U.S. patent application number 15/159076 was filed with the patent office on 2017-02-16 for imager arrays including an m x n array of focal planes in which different types of focal planes are distributed around a reference focal plane.
The applicant listed for this patent is Pelican Imaging Corporation. Invention is credited to Andrew Kenneth John McMahon, Bedabrata Pain.
Application Number | 20170048468 15/159076 |
Document ID | / |
Family ID | 44911485 |
Filed Date | 2017-02-16 |
United States Patent
Application |
20170048468 |
Kind Code |
A1 |
Pain; Bedabrata ; et
al. |
February 16, 2017 |
Imager Arrays Including an M x N Array of Focal Planes in which
Different Types of Focal Planes are Distributed Around a Reference
Focal Plane
Abstract
Architectures for imager arrays configured for use in array
cameras in accordance with embodiments of the invention are
described. One embodiment of the invention includes a plurality of
focal planes, where each focal plane comprises a two dimensional
arrangement of pixels having at least two pixels in each dimension
and each focal plane is contained within a region of the imager
array that does not contain pixels from another focal plane,
control circuitry configured to control the capture of image
information by the pixels within the focal planes, where the
control circuitry is configured so that the capture of image
information by the pixels in at least two of the focal planes is
separately controllable, and sampling circuitry configured to
convert pixel outputs into digital pixel data.
Inventors: |
Pain; Bedabrata; (Los
Angeles, CA) ; McMahon; Andrew Kenneth John; (Menlo
Park, CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Pelican Imaging Corporation |
Santa Clara |
CA |
US |
|
|
Family ID: |
44911485 |
Appl. No.: |
15/159076 |
Filed: |
May 19, 2016 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
13106797 |
May 12, 2011 |
|
|
|
15159076 |
|
|
|
|
61334011 |
May 12, 2010 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
H04N 5/23232 20130101;
H04N 9/09 20130101; H04N 5/3745 20130101; H04N 5/3415 20130101;
H04N 5/3653 20130101; H04N 9/045 20130101; H04N 5/3742 20130101;
H04N 5/378 20130101; H04N 9/04557 20180801; H04N 5/345 20130101;
H04N 5/341 20130101; H04N 5/36961 20180801 |
International
Class: |
H04N 5/341 20060101
H04N005/341; H04N 5/378 20060101 H04N005/378; H04N 9/09 20060101
H04N009/09 |
Claims
1. An imager array, comprising: a plurality of focal planes, where
each focal plane comprises a two dimensional arrangement of pixels
having at least two pixels in each dimension and each focal plane
is contained within a region of the imager array that does not
contain pixels from another focal plane; control circuitry
configured to control the capture of image information by the
pixels within the focal planes, where the control circuitry is
configured so that the capture of image information by the pixels
in at least two of the focal planes is separately controllable; and
sampling circuitry configured to convert pixel outputs into digital
pixel data.
2. The imager array of claim 1, wherein the plurality of focal
planes is arranged as a two dimensional array of focal planes
having at least three focal planes in one dimension.
3. The imager array of claim 1, wherein the plurality of focal
planes is arranged as a two dimensional array of focal planes
having at least three focal planes in both dimensions.
4. The imager array of claim 1, wherein the plurality of focal
planes arranged as an N.times.M array of focal planes comprising at
least two focal planes configured to capture blue light, at least
two focal planes configured to capture green light, and at least
two focal planes configured to capture red light.
5. The imager array of claim 1, wherein each focal plane comprises
rows and columns of pixels.
6. The imager array of claim 1, wherein the control circuitry is
configured to control capture of image information by a pixel by
controlling the resetting of the pixel.
7. The imager array of claim 1, wherein the control circuitry is
configured to control capture of image information by a pixel by
controlling the readout of the pixel.
8. The imager array of claim 1, wherein the control circuitry is
configured to control capture of image information by controlling
the integration time of each pixel.
9. The imager array of claim 1, wherein the control circuitry is
configured to control the processing of image information by
controlling the gain of the sampling circuitry.
10. The imager array of claim 1, wherein the control circuitry is
configured to control the processing of image information by
controlling the black level offset of each pixel.
11. The imager array of claim 1, wherein the control circuitry is
configured to control the capture of image information by
controlling readout direction.
12. The imager array of claim 11, wherein the read-out direction is
selected from the group consisting of: top to bottom; and bottom to
top.
13. The imager array of claim 11, wherein the read-out direction is
selected from the group consisting of: left to right; and right to
left.
14. The imager array of claim 1, wherein the control circuitry is
configured to control the capture of image information by
controlling the readout region of interest.
15. The imager array of claim 1, wherein the control circuitry is
configured to control the capture of image information by
controlling horizontal sub-sampling.
16. The imager array of claim 1, wherein the control circuitry is
configured to control the capture of image information by
controlling vertical sub-sampling.
17. The imager array of claim 1, wherein the control circuitry is
configured to control the capture of image information by
controlling pixel charge-binning.
18. The imager array of claim 1, wherein the imager array is a
monolithic integrated circuit imager array.
19. The imager array of claim 1, wherein a two dimensional array of
adjacent pixels in at least one focal plane have the same capture
band.
20. The imager array of claim 19, wherein the capture band is
selected from the group consisting of: blue light; cyan light;
extended color light comprising visible light and near-infra red
light; green light; infra-red light; magenta light; near-infra red
light; red light; yellow light; and white light.
Description
RELATED APPLICATION
[0001] This application is a continuation of U.S. patent
application Ser. No. 13/106,797, filed May 12, 2011, which claims
the benefit of U.S. Provisional Patent Application Ser. No.
61/334,011 filed on May 12, 2010. The disclosures of both U.S.
patent application Ser. No. 13/106,797 and U.S. Provisional Patent
Application Ser. No. 61/334,011 are incorporated by reference
herein in their entirety.
FIELD OF THE INVENTION
[0002] The present invention relates generally to imagers and more
specifically to imager arrays used in array cameras.
BACKGROUND OF THE INVENTION
[0003] A sensor used in a conventional single sensor camera,
typically includes a row controller and one or more column read-out
circuits. In the context of the array of pixels in an imager, the
term "row" is typically used to refer to a group of pixels that
share a common control line(s) and the term "column" is a group of
pixels that share a common read-out line(s). A number of array
camera designs have been proposed that use either an array of
individual cameras/sensors or a lens array focused on a single
focal plane sensor. When multiple separate cameras are used in the
implementation of an array camera, each camera has a separate I/O
path and the camera controllers are typically required to be
synchronized in some way. When a lens array focused on a single
focal plane sensor is used to implement an array camera, the sensor
is typically a conventional sensor similar to that used in a
conventional camera. As such, the sensor does not possess the
ability to independently control the pixels within the image circle
of each lens in the lens array.
SUMMARY OF THE INVENTION
[0004] Systems and methods are disclosed in which an imager array
is implemented as a monolithic integrated circuit in accordance
with embodiments of the invention. In many embodiments, the imager
array includes a plurality of imagers that are each independently
controlled by control logic within the imager array and the image
data captured by each imager is output from the imager array using
a common I/O path. In a number of embodiments, the pixels of each
imager are backside illuminated and the bulk silicon of the imager
array is thinned to different depths in the regions corresponding
to different imagers in accordance with the spectral wavelengths
sensed by each imager.
[0005] One embodiment of the invention includes a plurality of
focal planes, where each focal plane comprises a two dimensional
arrangement of pixels having at least two pixels in each dimension
and each focal plane is contained within a region of the imager
array that does not contain pixels from another focal plane,
control circuitry configured to control the capture of image
information by the pixels within the focal planes, where the
control circuitry is configured so that the capture of image
information by the pixels in at least two of the focal planes is
separately controllable, and sampling circuitry configured to
convert pixel outputs into digital pixel data.
[0006] In a further embodiment, the plurality of focal planes is
arranged as a two dimensional array of focal planes having at least
three focal planes in one dimension.
[0007] In another embodiment, the plurality of focal planes is
arranged as a two dimensional array of focal planes having at least
three focal planes in both dimensions.
[0008] In a still further embodiment, the plurality of focal planes
arranged as an N.times.M array of focal planes comprising at least
two focal planes configured to capture blue light, at least two
focal planes configured to capture green light, and at least two
focal planes configured to capture red light.
[0009] In still another embodiment, each focal plane comprises rows
and columns of pixels.
[0010] In a yet further embodiment, the control circuitry is
configured to control capture of image information by a pixel by
controlling the resetting of the pixel.
[0011] In yet another embodiment, the control circuitry is
configured to control capture of image information by a pixel by
controlling the readout of the pixel.
[0012] In a further embodiment again, the control circuitry is
configured to control capture of image information by controlling
the integration time of each pixel.
[0013] In another embodiment again, the control circuitry is
configured to control the processing of image information by
controlling the gain of the sampling circuitry.
[0014] In a further additional embodiment, the control circuitry is
configured to control the processing of image information by
controlling the black level offset of each pixel.
[0015] In another additional embodiment, the control circuitry is
configured to control the capture of image information by
controlling readout direction.
[0016] In a still yet further embodiment, the read-out direction is
selected from the group including top to bottom, and bottom to
top.
[0017] In still yet another embodiment, the read-out direction is
selected from the group including left to right, and right to
left.
[0018] In a still further embodiment again, the control circuitry
is configured to control the capture of image information by
controlling the readout region of interest.
[0019] In still another embodiment again, the control circuitry is
configured to control the capture of image information by
controlling horizontal sub-sampling.
[0020] In a still further additional embodiment, the control
circuitry is configured to control the capture of image information
by controlling vertical sub-sampling.
[0021] In still another additional embodiment, the control
circuitry is configured to control the capture of image information
by controlling pixel charge-binning.
[0022] In a yet further embodiment again, the imager array is a
monolithic integrated circuit imager array.
[0023] In yet another embodiment again, a two dimensional array of
adjacent pixels in at least one focal plane have the same capture
band.
[0024] In a yet further additional embodiment, the capture band is
selected from the group including blue light, cyan light, extended
color light comprising visible light and near-infra red light,
green light, infra-red light, magenta light, near-infra red light,
red light, yellow light, and white light.
[0025] In a further additional embodiment again, a first array of
adjacent pixels in a first focal plane have a first capture band, a
second array of adjacent pixels in a second focal plane have a
second capture band, where the first and second capture bands are
the same, the peripheral circuitry is configured so that the
integration time of the first array of adjacent pixels is a first
time period, and the peripheral circuitry is configured so that the
integration time of the second array of adjacent pixels is a second
time period, where the second time period is longer than the first
time period.
[0026] In another further embodiment, at least one of the focal
planes includes an array of adjacent pixels, where the pixels in
the array of adjacent pixels are configured to capture different
colors of light.
[0027] In yet another further embodiment, the array of adjacent
pixels employs a Bayer filter pattern.
[0028] In still another further embodiment, the plurality of focal
planes is arranged as a 2.times.2 array of focal planes, a first
focal plane in the array of focal planes includes an array of
adjacent pixels that employ a Bayer filter pattern, a second focal
plane in the array of focal planes includes an array of adjacent
pixels configured to capture green light, a third focal plane in
the array of focal planes includes an array of adjacent pixels
configured to capture red light, and a fourth focal plane in the
array of focal planes includes an array of adjacent pixels
configured to capture blue light.
[0029] In another further embodiment again, the plurality of focal
planes is arranged as a two dimensional array of focal planes
having at least three focal planes in one dimension.
[0030] In another further additional embodiment, the plurality of
focal planes is arranged as a two dimensional array of focal planes
having at least three focal planes in both dimensions.
[0031] In still yet another further embodiment, the control
circuitry comprises a global counter.
[0032] In still another further embodiment again, the control
circuitry is configured to stagger the start points of image
read-out such that each focal plane has a controlled temporal
offset with respect to a global counter.
[0033] In still another further additional embodiment again, the
control circuitry is configured to separately control the
integration times of the pixels in each focal plane based upon the
capture band of the pixels in the focal plane using the global
counter.
[0034] In yet another further embodiment again, the control
circuitry is configured to separately control the frame rate of
each focal plane based upon the global counter.
[0035] In yet another further additional embodiment, the control
circuitry further comprises a pair of pointers for each focal
plane.
[0036] In a still further embodiment, the offset between the
pointers specifies an integration time.
[0037] In still another embodiment, the offset between the pointers
is programmable.
[0038] In a yet further embodiment, the control circuitry comprises
a row controller dedicated to each focal plane.
[0039] In yet another embodiment, the imager array includes an
array of M.times.N focal planes, and the control circuitry
comprises a single row decoder circuit configured to address each
row of pixels in each row of M focal planes.
[0040] In a further embodiment again, the control circuitry is
configured to generate a first set of pixel level timing signals so
that the row decoder and a column circuit sample a first row of
pixels within a first focal plane, and the control circuitry is
configured to generate a second set of pixel level timing signals
so that the row decoder and a column circuit sample a second row of
pixels within a second focal plane.
[0041] In another embodiment again, each focal plane has dedicated
sampling circuitry.
[0042] In a further additional embodiment, at least a portion of
the sampling circuitry is shared by a plurality of the focal
planes.
[0043] In another additional embodiment, the imager array includes
an array of M.times.N focal planes, and the sampling circuitry
comprises M analog signal processors (ASPs) and each ASP is
configured to sample pixels read-out from N focal planes.
[0044] In a still yet further embodiment, each ASP is configured to
receive pixel output signals from the N focal planes via N inputs,
and each ASP is configured to sequentially process each pixel
output signal on its N inputs.
[0045] In still yet another embodiment, the control circuitry is
configured so that a single common analog pixel signal readout line
is shared by all pixels in a group of N focal planes, and the
control circuitry is configured to control the capture of image
data to time multiplex the pixel output signals received by each of
the M ASPs.
[0046] In still another embodiment again, the imager array includes
an array of M.times.N focal planes, the sampling circuitry
comprises a plurality of analog signal processors (ASPs) and each
ASP is configured to sample pixels read-out from a plurality of
focal planes, the control circuitry is configured so that a single
common analog pixel signal readout line is shared by all pixels in
the plurality of focal planes, and the control circuitry is
configured to control the capture of image data to time multiplex
the pixel output signals received by each of the plurality of
ASPs.
[0047] In a yet further embodiment again, the sampling circuitry
comprises analog front end (AFE) circuitry and analog-to-digital
conversion (ADC) circuitry.
[0048] In yet another embodiment again, the sampling circuitry is
configured so that each focal plane has a dedicated AFE and at
least one ADC is shared between at least two focal planes.
[0049] In a yet further additional embodiment, the sampling
circuitry is configured so that at least one ADC is shared between
a pair of focal planes.
[0050] In yet another additional embodiment, the sampling circuitry
is configured so that at least one ADC is shared between four focal
planes.
[0051] In a further additional embodiment again, the sampling
circuitry is configured so that at least one AFE is shared between
at least two focal planes.
[0052] In another additional embodiment again, the sampling
circuitry is configured so that at least one AFE is shared between
a pair of focal planes.
[0053] In another further embodiment, the sampling circuitry is
configured so that two pairs of focal planes that each share an AFE
collectively share an ADC.
[0054] In still another further embodiment, the control circuitry
is configured to separately control the power down state of each
focal plane and associated AFE circuitry or processing timeslot
therein.
[0055] In yet another further embodiment, the control circuitry
configures the pixels of at least one inactive focal plane to be in
a constant reset state.
[0056] In another further embodiment again, at least one focal
plane includes reference pixels to calibrate pixel data captured
using the focal plane.
[0057] In another further additional embodiment, the control
circuitry is configured to separately control the power down state
of the focal plane's associated AFE circuitry or processing
timeslot therein, and the control circuitry is configured to power
down the focal plane's associated AFE circuitry or processing
timeslot therein without powering down the associated AFE circuitry
or processing timeslot therein for readout of the reference pixels
of the focal plane.
[0058] In still yet another further additional embodiment, the
pixels in the array of adjacent pixels share read-out
circuitry.
[0059] In still another further embodiment again, the read-out
circuit includes a reset transistor, a floating diffusion
capacitor, and a source follower amplifier transistor.
[0060] In still another further additional embodiment, the array of
adjacent pixels in the at least one focal plane is a first array of
adjacent pixels, the imager array includes a second array of
adjacent pixels within another of the plurality of focal planes and
the pixels in the second array of adjacent pixels in the second
array of adjacent pixels, the capture band of the pixels in the
second array of adjacent pixels differs from the capture band of
the pixels in the first array of adjacent pixels, and the full well
capacity of the pixels in the first array of adjacent pixels is
different to the full well capacity of the pixels in the second
array of adjacent pixels.
[0061] In yet another further embodiment again, the full well
capacity of the pixels in the first array of adjacent pixels is
configured so that each pixel well is filled by the number of
electrons generated when the pixel is exposed for a predetermined
integration time to light within the first capture band having a
predetermined maximum spectral radiance, and the full well capacity
of the pixels in the second array of adjacent pixels is configured
so that each pixel well is filled by the number of electrons
generated when the pixel is exposed for a predetermined integration
time to light within the second capture band having a predetermined
maximum spectral radiance.
[0062] In yet another further additional embodiment, the floating
diffusion capacitance determines the conversion gain of each pixel
in the array of adjacent pixels.
[0063] In a further embodiment, the array of adjacent pixels in the
at least one focal plane is a first array of adjacent pixels, the
imager array includes a second array of adjacent pixels within
another of the plurality of focal planes and the pixels in the
second array of adjacent pixels has a floating diffusion
capacitance that determines the conversion gain of each pixel in
the second array of adjacent pixels, the capture band of the pixels
in the second array of adjacent pixels differs from the capture
band of the pixels in the first array of adjacent pixels, and the
conversion gain of the pixels in the first array of adjacent pixels
is different to the conversion gain of the pixels in the second
array of adjacent pixels.
[0064] In another embodiment, the floating diffusion capacitors of
the first and second arrays of adjacent pixels are configured to
minimize the input referred noise of the pixel outputs.
[0065] In a still further embodiment, the array of adjacent pixels
in the at least one focal plane is a first array of adjacent
pixels, the imager array includes a second array of adjacent pixels
within another of the plurality of focal planes and the pixels in
the second array of adjacent pixels has a floating diffusion
capacitance that determines the conversion gain of each pixel in
the second array of adjacent pixels, the capture band of the pixels
in the second array of adjacent pixels is the same as the capture
band of the pixels in the first array of adjacent pixels, and the
conversion gain of the pixels in the first array of adjacent pixels
is different to the conversion gain of the pixels in the second
array of adjacent pixels.
[0066] In still another embodiment, the source follower gain of
each pixel in the array determines the output voltage the
pixels.
[0067] In a yet further embodiment, the array of adjacent pixels in
the at least one focal plane is a first array of adjacent pixels,
the imager array includes a second array of adjacent pixels within
another of the plurality of focal planes and the pixels in the
second array of adjacent pixels has a fixed source follower gain
for each pixel in the second array of adjacent pixels, the capture
band of the pixels in the second array of adjacent pixels differs
from the capture band of the pixels in the first array of adjacent
pixels, and the source follower gain of the pixels in the first
array of adjacent pixels is different to the source follower gain
of the pixels in the second array of adjacent pixels.
[0068] In yet another embodiment, the source follower gain of the
first and second arrays of adjacent pixels are configured so that
the maximum output signal swing of each pixel is the same.
[0069] In a further embodiment again, a first array of adjacent
pixels in a first focal plane have a first capture band, a second
array of adjacent pixels in a second focal plane have a second
capture band, where the first and second capture bands differ, the
imager array is backside illuminated, and the thinning depth of the
imager array in the region containing the first array of adjacent
pixels is different to the thinning depth of the region of the
imager array containing the second array of adjacent pixels.
[0070] In another embodiment again, the first and second capture
bands do not overlap.
[0071] In a further additional embodiment, the thinning depth of
the imager in the array in the region containing the first array is
related to the first capture band, and the thinning depth of the
imager in the array in the region containing the second array is
related to the second capture band.
[0072] In another additional embodiment, the first thinning depth
is configured so as to position the peak carrier generation within
the photodiode's depletion region given a nominal capture band
wavelength of 450 nm.
[0073] In a still yet further embodiment, the first thinning depth
is configured so as to position the peak carrier generation within
the photodiode's depletion region given a nominal capture band
wavelength of 550 nm.
[0074] In still yet another embodiment, the first thinning depth is
configured so as to position the peak carrier generation within the
photodiode's depletion region given a nominal capture band
wavelength of 640nm.
[0075] In a still further embodiment again, a first array of
adjacent pixels in a first focal plane have a first capture band, a
second array of adjacent pixels in a second focal plane have a
second capture band, where the first and second capture bands
differ, the pixels in the first array of adjacent pixels are a
first pixel size, the pixels in the second array of adjacent pixels
are a second pixel size, and the first pixel size is larger than
the second pixel size and the first capture band includes longer
wavelengths of light than the second capture band.
[0076] In a still further additional embodiment, a first portion of
the control circuitry is located on one side of a focal plane and a
second portion of the control circuitry is located on the opposite
side of the focal plane.
[0077] In still another additional embodiment, the first portion of
the control circuitry is configured to control the capture of
information by a plurality of pixels in a first focal plane and in
plurality of pixels in a second focal plane located adjacent the
first focal plane.
[0078] In a yet further embodiment again, the imager array is
configured to receive a lens array positioned above the focal
planes of the imager array, and each of the plurality of focal
planes is located within a region in the imager array corresponding
to an image circle of the lens array, when a lens array is mounted
to the imager array.
[0079] Yet another embodiment again, also includes a cover-glass
mounted above the focal planes of the imager array.
[0080] In a further additional embodiment again, adjacent focal
planes are separated by a spacing distance.
[0081] In another additional embodiment again, control circuitry is
located within the spacing distance between adjacent focal
planes.
[0082] In another further embodiment, sampling circuitry is located
within the spacing distance between adjacent focal planes.
BRIEF DESCRIPTION OF THE DRAWINGS
[0083] FIG. 1 is a block diagram of an array camera in accordance
with an embodiment of the invention.
[0084] FIG. 1A is a block diagram of a monolithic imager array in
accordance with an embodiment of the invention.
[0085] FIGS. 2A-2B illustrate imager configurations of imager
arrays in accordance with embodiments of the invention.
[0086] FIG. 3 illustrates an architecture of an imager array in
accordance with an embodiment of the invention.
[0087] FIG. 4 illustrates another architecture of an imager array
including shared analog to digital converters in accordance with an
embodiment of the invention.
[0088] FIG. 4A illustrates a further architecture of an imager
array including shared column circuits in accordance with an
embodiment of the invention.
[0089] FIG. 4B illustrates still another architecture of an imager
array including shared split column circuits in accordance with an
embodiment of the invention.
[0090] FIG. 4C illustrates the phase shifting of column circuit
outputs from two focal planes read-out in accordance with an
embodiment of the invention.
[0091] FIG. 4D illustrates a pair of focal planes in an imager
array having dedicated analog front end circuitry and sharing an
analog to digital converter in accordance with an embodiment of the
invention.
[0092] FIG. 4E illustrates a group of four focal planes in an
imager array where pairs of focal planes share analog front end
circuitry and the group of four focal planes share an analog to
digital converter in accordance with an embodiment of the
invention.
[0093] FIG. 4F illustrates a pair of focal planes within an imager
array where the pair of focal planes share column control read-out
circuitry in accordance with an embodiment of the invention.
[0094] FIG. 4G illustrates a pair of focal planes within an imager
array where the column control and read-out circuitry is split and
a single block of column control and read-out circuitry reads out
odd columns from a first focal plane and even columns from a second
focal plane in accordance with an embodiment of the invention.
[0095] FIG. 4H is a block diagram illustrating focal plane timing
and control circuitry in accordance with an embodiment of the
invention.
[0096] FIG. 5 illustrates a backside illuminated imager array with
optimized thinning depths in accordance with an embodiment of the
invention.
DETAILED DISCLOSURE OF THE INVENTION
[0097] Turning now to the drawings, architectures for imager arrays
configured for use in array cameras in accordance with embodiments
of the invention are illustrated. In many embodiments, a
centralized controller on an imager array enables fine control of
the capture time of each focal plane in the array. The term focal
plane describes a two dimensional arrangement of pixels. Focal
planes in an imager array are typically non-overlapping (i.e. each
focal plane is located within a separate region on the imager
array). The term imager is used to describe the combination of a
focal plane and the control circuitry that controls the capture of
image information using the pixels within the focal plane. In a
number of embodiments, the focal planes of the imager array can be
separately triggered. In several embodiments, the focal planes of
the imager array utilize different integration times tailored to
the capture band of the pixels within each focal plane. The capture
band of a pixel typically refers to a contiguous sub-band of the
electromagnetic system to which a pixel is sensitive. In addition,
the specialization of specific focal planes so that all or a
majority of the pixels in the focal plane have the same capture
band enables a number of pixel performance improvements and
increases in the efficiency of utilization of peripheral circuitry
within the imager array.
[0098] In a number of embodiments, the pixels of the imager array
are backside illuminated and the substrate of the regions
containing each of the focal planes are thinned to different depths
depending upon the spectral wavelengths sensed by the pixels in
each focal plane. In addition, the pixels themselves can be
modified to improve the performance of the pixels with respect to
specific capture bands. In many embodiments, the conversion gain,
source follower gain and full well capacity of the pixels in each
focal plane are determined to improve the performance of the pixels
with respect to their specific capture bands.
[0099] In several embodiments, each focal plane possesses dedicated
peripheral circuitry to control the capture of image information.
In certain embodiments, the grouping of pixels intended to capture
the same capture band into focal planes enables peripheral
circuitry to be shared between the pixels. In many embodiments, the
analog front end, analog to digital converter, and/or column
read-out and control circuitry are shared between pixels within two
or more focal planes.
[0100] In many embodiments, the imagers in an imager array can be
placed in a lower power state to conserve power, which can be
useful in operating modes that do not require all imagers to be
used to generate the output image (e.g. lower resolution modes). In
several embodiments, the pixels of imagers in the low power state
are held with the transfer gate on so as to maintain the
photodiode's depletion region at its maximum potential and carrier
collection ability, thus minimizing the probability of photo
generated carriers generated in an inactive imager from migrating
to the pixels of active imagers. Array cameras and imager arrays in
accordance with embodiments of the invention are discussed further
below.
1. Array Camera Architecture
[0101] An array camera architecture that can be used in a variety
of array camera configurations in accordance with embodiments of
the invention is illustrated in FIG. 1. The array camera 100
includes an imager array 110, which is connected to an image
processing pipeline module 120 and to a controller 130.
[0102] The imager array 110 includes an M.times.N array of
individual and independent focal planes, each of which receives
light through a separate lens system. The imager array can also
include other circuitry to control the capture of image data using
the focal planes and one or more sensors to sense physical
parameters. The control circuitry can control imaging and
functional parameters such as exposure times, trigger times, gain,
and black level offset. The control circuitry can also control the
capture of image information by controlling read-out direction
(e.g. top-to-bottom or bottom-to-top, and left-to-right or
right-to-left). The control circuitry can also control read-out of
a region of interest, horizontal sub-sampling, vertical
sub-sampling, and/or charge-binning. In many embodiments, the
circuitry for controlling imaging parameters may trigger each focal
plane separately or in a synchronized manner. The imager array can
include a variety of other sensors, including but not limited to,
dark pixels to estimate dark current at the operating temperature.
Imager arrays that can be utilized in array cameras in accordance
with embodiments of the invention are disclosed in PCT Publication
WO 2009/151903 to Venkataraman et al., the disclosure of which is
incorporated herein by reference in its entirety. In a monolithic
implementation, the imager array may be implemented using a
monolithic integrated circuit. When an imager array in accordance
with embodiments of the invention is implemented in a single
self-contained SOC chip or die, the imager array can be referred to
as an imager array. The term imager array can be used to describe a
semiconductor chip on which the imager array and associated
control, support, and read-out electronics are integrated.
[0103] The image processing pipeline module 120 is hardware,
firmware, software, or a combination thereof for processing the
images received from the imager array 110. The image processing
pipeline module 120 typically processes the multiple low resolution
(LR) images captured by the camera array and produces a synthesized
higher resolution image in accordance with an embodiment of the
invention. In a number of embodiments, the image processing
pipeline module 120 provides the synthesized image data via an
output 122. Various image processing pipeline modules that can be
utilized in a camera array in accordance with embodiments of the
invention are disclosed in U.S. patent application Ser. No.
12/967,807 entitled "System and Methods for Synthesizing High
Resolution Images Using Super-Resolution Processes" filed Dec. 14,
2010, the disclosure of which is incorporated by reference herein
in its entirety.
[0104] The controller 130 is hardware, software, firmware, or a
combination thereof for controlling various operation parameters of
the imager array 110. In many embodiments, the controller 130
receives inputs 132 from a user or other external components and
sends operation signals to control the imager array 110. The
controller 130 can also send information to the image processing
pipeline module 120 to assist processing of the LR images captured
by the imager array 110.
[0105] Although a specific array camera architecture is illustrated
in FIG. 1, alternative architectures for capturing a plurality of
images of a scene using an imager array can also be utilized in
accordance with embodiments of the invention. Operation of array
cameras, imager array configurations, and processing multiple
captured images of a scene in accordance with embodiments of the
invention are discussed further below.
2. Imager Array Architectures
[0106] An imager array in accordance with an embodiment of the
invention is illustrated in FIG. 1A. The imager array includes a
focal plane array core 152 that includes an array of focal planes
153 and all analog signal processing, pixel level control logic,
signaling, and analog-to-digital conversion circuitry. The imager
array also includes focal plane timing and control circuitry 154
that is responsible for controlling the capture of image
information using the pixels. In a number of embodiments, the focal
plane timing and control circuitry utilizes reset and read-out
signals to control the integration time of the pixels. In other
embodiments, any of a variety of techniques can be utilized to
control integration time of pixels and/or to capture image
information using pixels. In many embodiments, the focal plane
timing and control circuitry 154 provides flexibility of image
information capture control, which enables features including (but
not limited to) high dynamic range imaging, high speed video, and
electronic image stabilization. In various embodiments, the imager
array includes power management and bias generation circuitry 156.
The power management and bias generation circuitry 156 provides
current and voltage references to analog circuitry such as the
reference voltages against which an ADC would measure the signal to
be converted against. In many embodiments, the power management and
bias circuitry also includes logic that turns off the
current/voltage references to certain circuits when they are not in
use for power saving reasons. In several embodiments, the imager
array includes dark current and fixed pattern (FPN) correction
circuitry 158 that increases the consistency of the black level of
the image data captured by the imager array and can reduce the
appearance of row temporal noise and column fixed pattern noise. In
several embodiments, each focal plane includes reference pixels for
the purpose of calibrating the dark current and FPN of the focal
plane and the control circuitry can keep the reference pixels
active when the rest of the pixels of the focal plane are powered
down in order to increase the speed with which the imager array can
be powered up by reducing the need for calibration of dark current
and FPN. In many embodiments, the SOC imager includes focal plane
framing circuitry 160 that packages the data captured from the
focal planes into a container file and can prepare the captured
image data for transmission. In several embodiments, the focal
plane framing circuitry includes information identifying the focal
plane and/or group of pixels from which the captured image data
originated. In a number of embodiments, the imager array also
includes an interface for transmission of captured image data to
external devices. In the illustrated embodiment, the interface is a
MIPI CSI 2 output interface supporting four lanes that can support
read-out of video at 30 fps from the imager array and incorporating
data output interface circuitry 162, interface control circuitry
164 and interface input circuitry 166. Typically, the bandwidth of
each lane is optimized for the total number of pixels in the imager
array and the desired frame rate. The use of various interfaces
including the MIPI CSI 2 interface to transmit image data captured
by an array of imagers within an imager array to an external device
in accordance with embodiments of the invention is described in
U.S. Provisional Patent Application No. 61/484,920, entitled
"Systems and Methods for Transmitting Array Camera Data", filed May
11, 2011, the disclosure of which is incorporated by reference
herein in its entirety. Although specific components of an imager
array architecture are discussed above with respect to FIG. 1A. As
is discussed further below, any of a variety of imager arrays can
be constructed in accordance with embodiments of the invention that
enable the capture of images of a scene at a plurality of focal
planes in accordance with embodiments of the invention.
Accordingly, focal plane array cores and various components that
can be included in imager arrays in accordance with embodiments of
the invention are discussed further below.
3. Focal Plane Array Cores
[0107] Focal plan array cores in accordance with embodiments of the
invention include an array of imagers and dedicated peripheral
circuitry for capturing image data using the pixels in each focal
plane. Imager arrays in accordance with embodiments of the
invention can include focal plan array cores that are configured in
any of a variety of different configurations appropriate to a
specific application. For example, customizations can be made to a
specific imager array designs including (but not limited to) with
respect to the focal plane, the pixels, and the dedicated
peripheral circuitry. Various focal plane, pixel designs, and
peripheral circuitry that can be incorporated into focal plane
array cores in accordance with embodiments of the invention are
discussed below.
3.1. Formation of Focal Planes on an Imager Array
[0108] An imager array can be constructed in which the focal planes
are formed from an array of pixel elements, where each focal plane
is a sub-array of pixels. In embodiments where each sub-array has
the same number of pixels, the imager array includes a total of
K.times.L pixel elements, which are segmented in M.times.N
sub-arrays of X.times.Y pixels, such that K=M.times.X, and
L=N.times.Y. In the context of an imager array, each sub-array or
focal plane can be used to generate a separate image of the scene.
Each sub-array of pixels provides the same function as the pixels
of a conventional imager (i.e. the imager in a camera that includes
a single focal plane).
[0109] As is discussed further below, an imager array in accordance
with embodiments of the invention can include a single controller
that can separately sequence and control each focal plane. Having a
common controller and I/O circuitry can provide important system
advantages including lowering the cost of the system due to the use
of less silicon area, decreasing power consumption due to resource
sharing and reduced system interconnects, simpler system
integration due to the host system only communicating with a single
controller rather than M.times.N controllers and read-out I/O
paths, simpler array synchronization due to the use of a common
controller, and improved system reliability due to the reduction in
the number of interconnects.
3.2. Layout of Imagers
[0110] As is disclosed in P.C.T. Publication WO 2009/151903
(incorporated by reference above), an imager array can include any
N.times.M array of focal planes such as the imager array (200)
illustrated in FIG. 2A. Each of the focal planes typically has an
associated filter and/or optical elements and can image different
wavelengths of light. In a number of embodiments, the imager array
includes focal planes that sense red light (R), focal planes that
sense green light (G), and focal planes that sense blue light (B).
Although in several embodiments, one or more of the focal planes
include pixels that are configured to capture different colors of
light. In a number of embodiments, the pixels employ a Bayer filter
pattern (or similar) pattern that enables different pixels within a
focal plane to capture different colors of light. In several
embodiments, a 2.times.2 imager array can include a focal plane
where the pixels employ a Bayer filter pattern (or similar), a
focal plane where the pixels are configured to capture blue light,
a focal plane where the image is configured to capture green light,
and a focal plane where the imager is configured to capture red
light. Array cameras incorporating such sensor arrays can utilize
the color information captured by the blue, green, and red focal
planes to enhance the colors of the image captured using the focal
plane that employs the Bayer filter. In other embodiments, the
focal plane that employs the Bayer pattern is incorporated into an
imager array that includes a two dimensional arrangement of focal
planes where there are at least three focal planes in one of the
dimensions. In a number of embodiments, there are at least three
focal planes in both dimensions.
[0111] The human eye is more sensitive to green light than to red
and blue light, therefore, an increase in the resolution of an
image synthesized from the low resolution image data captured by an
imager array can be achieved using an array that includes more
focal planes that sense green light than focal planes that sense
red or blue light. A 5.times.5 imager array (210) including 17
focal planes that sense green light (G), four focal planes that
sense red light (R), and four focal planes that sense blue light
(B) is illustrated in FIG. 2B. In several embodiments, the imager
array also includes focal planes that sense near-IR wavelengths or
extended-color wavelengths (i.e. spanning both color and near-IR
wavelengths), which can be used to improve the performance of the
array camera in low light conditions. In other embodiments, the
5.times.5 imager array includes at least 13 focal planes, at least
15 focal planes or at least 17 focal planes. In addition, the
5.times.5 imager array can include at least four focal planes that
sense red light, and/or at least four focal planes that sense blue
light. In addition, the number of focal planes that sense red light
and the number of focal planes that sense blue light can be the
same, but need not be the same. Indeed, several imager arrays in
accordance with embodiments of the invention include different
numbers of focal planes that sense red light and that sense blue
light. In many embodiments, other arrays are utilized including
(but not limited to) 3.times.2 arrays, 3.times.3 arrays, 3.times.4
arrays, 4.times.4 arrays, 4.times.5 arrays, 4.times.6 arrays,
5.times.5 arrays, 5.times.6 arrays, 6.times.6 arrays, and 3.times.7
arrays. In a number of embodiments, the imager array includes a two
dimensional array of focal planes having at least three focal
planes in one of the dimensions. In several embodiments, there are
at least three focal plane in both dimensions of the array. In
several embodiments, the array includes at least two focal planes
having pixels configured to capture blue light, at least two focal
planes having pixels configured to capture green light, and at
least two focal planes having pixels configured to capture red
light.
[0112] Additional imager array configurations are disclosed in U.S.
patent application Ser. No. 12/952,106 entitled "Capturing and
Process of Images Using Monolithic Camera Array with Heterogeneous
Imagers" to Venkataraman et al., the disclosure of which is
incorporated by reference herein in its entirety.
[0113] Although specific imager array configurations are disclosed
above, any of a variety of regular or irregular layouts of imagers
including imagers that sense visible light, portions of the visible
light spectrum, near-IR light, other portions of the spectrum
and/or combinations of different portions of the spectrum can be
utilized to capture images that provide one or more channels of
information for use in SR processes in accordance with embodiments
of the invention. The construction of the pixels of an imager in an
imager array in accordance with an embodiment of the invention can
depend upon the specific portions of the spectrum imaged by the
imager. Different types of pixels that can be used in the focal
planes of an imager array in accordance with embodiments of the
invention are discussed below.
3.3. Pixel Design
[0114] Within an imager array that is designed for color or
multi-spectral capture, each individual focal plane can be
designated to capture a sub-band of the visible spectrum. Each
focal plane can be optimized in various ways in accordance with
embodiments of the invention based on the spectral band it is
designated to capture. These optimizations are difficult to perform
in a legacy Bayer pattern based image sensor since the pixels
capturing their respective sub-band of the visible spectrum are all
interleaved within the same pixel array. In many embodiments of the
invention, backside illumination is used where the imager array is
thinned to different depths depending upon the capture band of a
specific focal plane. In a number of embodiments, the sizes of the
pixels in the imager array are determined based upon the capture
band of the specific imager. In several embodiments, the conversion
gains, source follower gains, and full well capacities of groups of
pixels within a focal plane are determined based upon the capture
band of the pixels. The various ways in which pixels can vary
between focal planes in an imager array depending upon the capture
band of the pixel are discussed further below.
3.3.1. Backside Illuminated Imager Array with Optimized Thinning
Depths
[0115] A traditional image sensor is illuminated from the front
side where photons must first travel through a dielectric stack
before finally arriving at the photodiode, which lies at the bottom
of the dielectric stack in the silicon substrate. The dielectric
stack exists to support metal interconnects within the device.
Front side illumination suffers from intrinsically poor Quantum
Efficiency (QE) performance (the ratio of generated carriers to
incident photons), due to problems such as the light being blocked
by metal structures within the pixel. Improvement is typically
achieved through the deposition of micro-lens elements on top of
the dielectric stack for each pixel so as to focus the incoming
light in a cone that attempts to avoid the metal structures within
the pixel.
[0116] Backside illumination is a technique employed in image
sensor fabrication so as to improve the QE performance of imagers.
In backside illumination (BSI), the silicon substrate bulk is
thinned (usually with a chemical etch process) to allow photons to
reach the depletion region of the photodiode through the backside
of the silicon substrate. When light is incident on the backside of
the substrate, the problem of aperturing by metal structures
inherent in frontside illumination is avoided. However, the
absorption depth of light in silicon is proportional to the
wavelength such that the red photons penetrate much deeper than
blue photons. If the thinning process does not remove sufficient
silicon, the depletion region will be too deep to collect photo
electrons generated from blue photons. If the thinning process
removes too much silicon, the depletion region can be too shallow
and red photons may travel straight though without interacting and
generating carriers. Red photons could also be reflected from the
front surface back and interact with incoming photons to create
constructive and destructive interference due to minor differences
in the thickness of the device. The effects caused by variations in
the thickness of the device can be evident as fringing patterns
and/or as spiky spectral QE response.
[0117] In a conventional imager, a mosaic of color filters
(typically a Bayer filter) is often used to provide RGB color
capture. When a mosaic based color imager is thinned for BSI, the
thinning depth is typically the same for all pixels since the
processes used do not thin individual pixels to different depths.
The common thinning depth of the pixels results in a necessary
balancing of QE performance between blue wavelengths and
red/near-IR wavelengths. An imager array in accordance with
embodiments of the invention includes an array of imagers, where
each pixel in a focal plane senses the same spectral wavelengths.
Different focal planes can sense different sub-bands of the visible
spectrum or indeed any sub-band of the electromagnetic spectrum for
which the band-gap energy of silicon has a quantum yield gain
greater than 0. Therefore, performance of an imager array can be
improved by using BSI where the thinning depth for the pixels of a
focal plane is chosen to match optimally the absorption depth
corresponding to the wavelengths of light each pixel is designed to
capture. In a number of embodiments, the silicon bulk material of
the imager array is thinned to different thicknesses to match the
absorption depth of each camera's capture band within the depletion
region of the photodiode so as to maximize the QE.
[0118] An imager array in which the silicon substrate is thinned to
different depths in regions corresponding to focal planes (i.e.
sub-arrays) that sense different spectral bandwidths in accordance
with an embodiment of the invention is conceptually illustrated in
FIG. 5. The imager array 500 includes a silicon substrate 502 on
the front side of which a dielectric stack and metal interconnects
504 are formed. In the illustrated embodiment, the silicon
substrate includes regions 506, 508, 510 in which the photodiodes
of pixels forming a focal plane for sensing blue light, the
photodiodes of pixels forming a focal plane for sensing green
light, and the photodiodes of pixels forming a focal plane for
sensing red light respectively are located. The backside of the
silicon substrate is thinned to different depths in each region. In
the illustrated embodiment, the substrate is thinned to correspond
to the absorption depth of 450 nm wavelength light (i.e.
approximately 0.4 .mu.m) in the region 506 in which the photodiodes
of pixels forming an imager for sensing blue light are located, the
substrate is thinned to correspond to the absorption depth of 550
nm wavelength light (i.e. approximately 1.5 .mu.m) in the region
508 in which the photodiodes of pixels forming an imager for
sensing green light are located, and the substrate is thinned to
correspond to the absorption depth of 640 nm wavelength light (i.e.
approximately 3.0 .mu.m) in the region 510 in which the photodiodes
of pixels forming an imager for sensing red light are located.
Although specific depths are shown in FIG. 5, other depths
appropriate to the spectral wavelengths sensed by a specific imager
and the requirements of the application can be utilized in
accordance with embodiments of the invention. In addition,
different thinning depths can also be used in array cameras that
are not implemented using imager arrays in accordance with
embodiments of the invention.
[0119] In many embodiments, the designation of color channels to
each imager within the array is achieved via a first filtration of
the incoming photons through a band-pass filter within the optical
path of the photons to the photodiodes. In several embodiments, the
thinning depth itself is used to create the designation of capture
wavelengths since the depletion region depth defines the spectral
QE of each imager.
3.3.2. Optimization of Pixel Size
[0120] Additional SNR benefits can be achieved by changing the
pixel sizes used in the imagers designated to capture each sub-band
of the spectrum. As pixel sizes shrink, the effective QE of the
pixel decreases since the ratio of photodiode depletion region area
to pixel area decreases. Microlenses are typically used to attempt
to compensate for this and they become more important as the pixel
size shrinks. Another detriment to pixel performance by pixel size
reduction comes from increased noise. To attempt to maintain the
balance of photo-active to read-out circuit area, in many
embodiments, the pixel transfer gate, source follower amplifier
transistor and reset transistors are also made smaller. As these
transistors reduce in size, numerous performance parameters are
degraded typically resulting in noise increase.
[0121] Electrical "cross-talk" also increases as a function of
reduced pixel-to-pixel spacing. Long wavelength photons penetrate
deeper into the substrate before interacting with the silicon to
create a charge carrier. These charge carriers wander in a somewhat
random fashion before resurfacing and collection in a photodiode
depletion region. This "circle" of probable resurface and
collection increases as a function of generation depth. Thus the
smaller the pixels become, the greater the number of pixels the
circle of probable resurface covers. This effect results in a
degradation of the Modulation Transfer Function (MTF) with increase
in photon wavelength.
[0122] Imagers designated to capture longer wavelengths can
therefore be optimized to improve system SNR by increasing the
pixel size and thus increasing the QE of the pixel. Since MTF drops
as a function of increased wavelength, the benefit of smaller
pixels for resolution purposes is diminished with increased
wavelength. Overall system resolution can thus be maintained while
increasing the pixel size for longer wavelengths so as to improve
QE and thus improve the overall system SNR. Although in many
embodiments, imager arrays in accordance with embodiments of the
invention utilize as small pixels as can be manufactured.
Accordingly, increasing pixel size in the manner outlined above is
simply one technique that can be utilized to improve camera
performance and the specific pixel size chosen typically depends
upon the specific application.
3.3.3. Imager Optimization
[0123] The push for smaller and smaller pixels has encouraged pixel
designers to re-architect the pixels such that they share read-out
circuits within a neighborhood. For example, a group of four
photodiodes may share the same reset transistor, floating diffusion
node and source follower amplifier transistors. When the four
pixels are arranged in a Bayer pattern arrangement, the group of
four pixels covers the full visible spectrum of capture. In imager
arrays in accordance with embodiments of the invention, these
shared pixel structures can be adapted to tailor the performance of
pixels in a focal plane to a given capture band. The fact that
these structures are shared by pixels that have different capture
bands in a traditional color filter array based image sensor means
that the same techniques for achieving performance improvements are
typically not feasible. The improvement of the performance of
pixels in a focal plane by selection of conversion gain, source
follower gain, and full well capacity based upon the capture band
of the pixels is discussed below. Although the discussion that
follows is with reference to 4T CMOS pixels, similar improvements
to pixel performance can be achieved in any imager array in which
pixels share circuitry in accordance with embodiments of the
invention.
3.3.3.1. Optimization of Conversion Gain
[0124] The performance of imagers within an imager array that are
intended to capture specific sub-bands of the spectrum can be
improved by utilizing pixels with different conversion gains
tailored for each of the different capture bands. Conversion gain
in a typical 4T CMOS pixel can be controlled by changing the size
of the capacitance of the "sense node", typically a floating
diffusion capacitor (FD). The charge to voltage conversion follows
the equation V=Q/C where Q is the charge, C is the capacitance and
V is the voltage. Thus the smaller the capacitance, the higher the
voltage resulting from a given charge hence the higher the
charge-to-voltage conversion gain of the pixel. The conversion gain
cannot obviously be increased infinitely however. The apparent full
well capacity of the pixel (number of photo-electrons the pixel can
record) will decrease if the capacitance of the FD becomes too
small. This is because the electrons from the photodiode transfer
into the FD due to a potential difference acting on them. Charge
transfer will stop when the potential difference is zero (or a
potential barrier exists between the PF and the FD). Thus if the
capacitance of the FD is too small, the potential equilibrium may
be reached before all electrons have been transferred out of the
photodiode.
3.3.3.2. Optimization of Source Follower Gain
[0125] Additional performance gains can be achieved by changing the
characteristics of the amplifiers in each pixel within a focal
plane. The amplifier in a traditional 4T CMOS pixel is constructed
from a Source Follower transistor. The Source Follower transistor
amplifies the voltage across the FD so as to drive the pixel signal
down the column line to the column circuit where the signal is
subsequently sampled.
[0126] The output voltage swing as a function of the input voltage
swing (i.e. the Source Follower amplifier's gain) can be controlled
during fabrication by changing the implant doping levels. Given the
pixel photodiode's full well capacity (in electrons) and the
capacitance of the FD, a range of voltages are established at the
input of the Source Follower transistor by the relationship
Vin=Vrst-Q/C where Vrst is the reset voltage of the FD, Q is the
charge of the electrons transferred to the FD from the photodiode
and C is the capacitance of the FD.
[0127] The photodiode is a pinned structure such that the range of
charge that may be accumulated is between 0 electrons and the full
well capacity. Therefore, with a given full well capacity of the
photodiode and a given capacitance of the FD and a desired output
signal swing of the source follower, the optimal gain or a near
optimal gain for the source follower transistor can be
selected.
3.3.3.3. Optimization of Full Well Capacity
[0128] Another optimization that can be performed is through
changing the full well capacity of the photodiodes. The full well
capacity of the photodiode is the maximum number of electrons the
photodiode can store in its maximally depleted state. The full well
of the pixels can be controlled through the x-y size of the
photodiode, the doping levels of the implants that form the diode
structure and the voltage used to reset the pixel.
3.3.3.4. Three Parameter Optimization
[0129] As can be seen in the previous sections, there are three
main characteristics that can be tuned in order to configure pixels
within a focal plane that have the same capture band for improved
imaging performance. The optimal solution for all three parameters
is dependent on the targeted behavior of a particular focal plane.
Each focal plane can be tailored to the spectral band it is
configured to capture. While the design of the pixel can be
optimized, in many embodiments the performance of the pixels is
simply improved with respect to a specific capture band (even
though the improvement may not be optimal). An example optimization
is as follows and similar processes can be used to simply improve
the performance of a pixel with respect to a specific capture
band:
a. Optimization of the Photodiode Full Well Capacity.
[0130] Given the speed of the optics and the transmittance of the
color filters, it is possible to estimate the number of electrons
that will be generated given a minimum integration time (e.g. 50
.mu.s) for a given maximum spectral radiance. Each sub-band of the
spectrum (color) will likely have a different number of electrons
generated. The full well capacities of the photodiodes for each
sub-band (color) can be chosen such that the maximum radiance
within that band under minimum integration times will fill the
well. The means by which this target full well capacity is achieved
could be through changing the x-y dimensions, changing the doping
levels during diode fabrication, changing the reset voltage of the
pixels or a combination of two or more of these parameters.
b. Optimization of Conversion Gain
[0131] The next step is to optimize the conversion gain of the
pixels. Given the number of electrons defined in the full well
optimization step, an optimal capacitance for the floating
diffusion can be chosen. The optimal capacitance is one, which
maintains a potential difference to support charge transfer from
the FD such that the full well capacity can be transferred in a
reasonable duration of time. The goal of this optimization is to
choose the smallest capacitance possible such that the charge to
voltage conversion gain is as high as possible such that input
referred noise is minimized and hence the maximum SNR for each
color channel is realized.
c. Optimization of Source Follower Gain
[0132] Once the optimal full-well capacity and charge to voltage
conversion gain is determined, the source follower amplifier gain
can be chosen. The difference between the reset voltage of the FD
(Vrst) and the voltage of the FD containing a full well charge load
(Vrst-Q/C) enables the definition of an optimal gain for the source
follower amplifier. The source follower gain defines the output
signal swing between Vrst and Vrst-Q/C. The optimal signal swing is
defined by such parameters as the operating voltage of the analog
signal processing and the A/D converter that sample and covert the
pixel output signal. The source follower gain is chosen for each
color channel such that their respective signal swings are all
matched to each other and match the maximum signal swing supported
by the analog signal processing and A/D converter circuits.
[0133] Having performed these pixel level optimizations on a per
capture band basis, the system will have the maximum SNR and
dynamic range for each capture band given linear operation.
Although the process described above is designed to provide an
optimal solution with regard to maximum SNR and dynamic range,
other design criteria can be used in the selection of the three
parameters described above to provide improved pixel performance
with respect to a specific capture band or application specific
desired behavior.
3.3.4. Dynamic Range Tailoring
[0134] Further optimizations of imager arrays can be achieved by
using pixels of different conversion gains within the same spectral
band. For example, the "green" imagers could be constructed from
pixels that have two or more different conversion gains. Therefore,
each "green" imager includes pixels that have a homogeneous
conversion gain, which is different to the conversion gain of
pixels in another of the "green" imagers in the array.
Alternatively, each imager could be constructed from a mosaic of
pixels having different conversion gains.
[0135] As mentioned previously, as the conversion gain increases
beyond a certain threshold, the input referred noise continues to
decrease but at the expense of effective full well capacity. This
effect can be exploited to yield a system having a higher dynamic
range. For example, half of all "green" focal planes could be
constructed using a conversion gain that optimizes both input
referred noise and full well capacity (a "normal green"). The other
half of all "green" focal planes could be constructed from pixels
that have a higher conversion gain, hence lower input referred
noise and lower effective full well capacity ("fast green"). Areas
of a scene having a lower light level could be recovered from the
"fast green" pixels (that are not saturated) and areas of brighter
light level could be recovered from the "normal green" pixels. The
result is an overall increase in dynamic range of the system.
Although, a specific 50/50 allocation of focal planes between "fast
green" and "normal green" is discussed above the number of focal
planes dedicated to "fast" imaging and the number of focal planes
dedicated to "normal" imaging is entirely dependent upon the
requirements of a specific application. In addition, separate focal
planes dedicated to "fast" and "normal" imaging can be utilized to
increase the dynamic range of other spectral bands and is not
simply limited to increasing the dynamic range with which an imager
array captures green light.
[0136] A similar effect could be achieved by controlling the
integration time of the "fast" and "normal" green sub-arrays such
that the "fast" pixels integrate for longer. However in a
non-stationary scene, this could result in motion artifacts since
the "fast" pixels would integrate the scene motion for longer than
the "normal" pixels creating an apparent spatial disparity between
the two green channels, which may be undesirable.
[0137] [Another alternative embodiments is that there are
additional AFE processing channels per focal planes that provide
different gains to different sets of pixels to achieve high dynamic
range. Another embodiment would be to have the amplifier change its
gain as the pixels are going through. We need another FIG. 4l to
show this with AFE at each end. Basically doing this with analog
signal processing after the charge to voltage conversion--In many
embodiments, the noise in the amplifier in the analog front end
does not vary considerably with gain. Therefore, increasing the
gain of the analog front end reduces the relative signal to noise
ratio of the amplified signal. Step size of the digitizer
inherently adds a quantization noise signal that you will amplify
when you try and scale the signal in the digital domain. Therefore,
the resulting amplified signal will have a lower SNR when amplified
prior to digitization using the analog amplifier than amplifying
both the signal after the signal has been digitized.
[0138] [Information Disclosure Statement--U.S. Pat. No.
6,774,941]
[0139] Although the present invention has been described in certain
specific embodiments, many additional modifications and variations
would be apparent to those skilled in the art. It is therefore to
be understood that the present invention may be practiced otherwise
than specifically described, including various changes in the size,
shape and materials, without departing from the scope and spirit of
the present invention. Thus, embodiments of the present invention
should be considered in all respects as illustrative and not
restrictive.
3.4. Peripheral Circuitry
[0140] In a conventional imager, pixels are typically accessed in a
row-wise fashion using horizontal control lines that run across
each row of pixels. Output signal lines that run vertically through
each pixel are used to connect the pixel output to a sampling
circuit at the column periphery. The horizontal control lines and
the output signal lines are typically implemented as metal traces
on silicon. The outputs from all pixels in a row are simultaneously
sampled at the column periphery, and scanned out sequentially using
column controllers. However, common row-wise access along the full
row of K pixels in an imager array does not enable the imagers to
be read out independently. As noted above, many of the benefits of
utilizing an imager array derive from the independence of the focal
planes and the ability for the imager array to separately control
the capture of image information by the pixels in each focal plane.
The ability to separately control the capture of information means
that the capture of image information by the pixels in a focal
plane can be customized to the spectral band the focal plane is
configured to capture. In a number of embodiments, the ability to
provide separate trigger times can be useful in synchronizing the
capture of images using focal planes that have different
integration times and in capturing sequences of images that can be
registered to provide slow motion video sequences. In order to
control the capture of image information by different focal planes
within an imager array, independent read-out control can be
provided for each focal plane. In several embodiments, the imager
array has independent read-out control due to the fact that each
focal plane has an associated row (column) controller, column (row)
read-out circuits and a dedicated pixel signal analog processor and
digitizer. In many embodiments, separate control of the capture of
image information by pixels in different focal planes is achieved
using peripheral circuitry that is shared between focal planes.
Imager arrays implemented using dedicated peripheral circuitry and
shared peripheral circuitry in accordance with embodiments of the
invention are discussed below.
3.4.1. Dedicated Peripheral Circuitry
[0141] An imager array including multiple focal planes having
independent read-out control and pixel digitization, where each
focal plane has dedicated peripheral circuitry, in accordance with
embodiments of the invention is illustrated in FIG. 3. The imager
array 300 includes a plurality of sub-arrays of pixels or focal
planes 302. Each focal plane has dedicated row control logic
circuitry 304 at its periphery, which is controlled by a common row
timing control logic circuitry 306. Although the column circuits
and row decoder are shown as a single block on one side of the
focal plane, the depiction as a single block is purely conceptual
and each logic block can be split between the left/right and/or
top/bottom of the focal plane so as to enable layout at double the
pixel pitch. Laying out the control and read-out circuitry in this
manner can result in a configuration where even columns are sampled
in one bank of column (row) circuits and odd columns would be
sampled in the other.
[0142] In a device including M.times.N focal planes, the read-out
control logic includes M sets of column control outputs per row of
focal planes (N). Each column sampling/read-out circuit 308 can
also have dedicated sampling circuitry for converting the captured
image information into digital pixel data. In many embodiments, the
sampling circuitry includes Analog Signal Processor (ASP), which
includes an Analog Front End (AFE) amplifier circuit and an Analog
to Digital Converter (ADC) 310. In other embodiments, any of a
variety of analog circuitry can be utilized to convert captured
image information into digitized pixel information. An ASP can be
implemented in a number of ways, including but not limited to, as a
single ASP operating at X pixel conversion per row period, where X
is the number of pixels in a row of the focal plane served by the
column sampling circuit (e.g. with a pipe-lined or SAR ADC), as X
ASPs operating in parallel at 1 pixel conversion per row period or
P ASPs operating in parallel at X/P conversions per row (see
discussion below). A common read-out control circuit 312 controls
the read-out of the columns in each imager.
[0143] In the illustrated embodiment, the master control logic
circuitry 314 controls the independent read-out of each imager. The
master control logic circuitry 314 includes high level timing
control logic circuitry to control the image capture and read-out
process of the individual focal plane. In a number of embodiments,
the master control portion of this block can implement features
including but not limited to: staggering the start points of image
read-out such that each focal plane has a controlled temporal
offset with respect to a global reference; controlling integration
times of the pixels within specific focal planes to provide
integration times specific to the spectral bandwidths being imaged;
the horizontal and vertical read-out direction of each imager; the
horizontal and vertical sub-sampling/binning/windowing of the
pixels within each focal plane; the frame/row/pixel rate of each
focal plane; and the power-down state control of each focal
plane.
[0144] The master control logic circuitry 314 handles collection of
pixel data from each of the imagers. In a number of embodiments,
the master control logic circuitry packs the image data into a
structured output format. Given that fewer than M.times.N output
ports are used to output the image data (e.g. there are 2 output
ports), the imager data is time multiplexed onto these output
ports. In a number of embodiments, a small amount of memory (FIFO)
is used to buffer the data from the pixels of the imagers until the
next available time-slot on the output port 316 and the master
control logic circuitry 314 or other circuitry in the imager array
periodically inserts codes into the data stream providing
information including, but not limited to, information identifying
a focal plane, information identifying a row and/or column within a
focal plane, and/or information identifying the relative time at
which the capture or read-out process began/ended for one or more
of the focal planes. Relative time information can be derived from
an on-chip timer or counter, whose instantaneous value can be
captured at the start/end of read-out of the pixels from each
imager either at a frame rate or a line rate. Additional codes can
also be added to the data output so as to indicate operating
parameters such as (but not limited to) the integration time of
each focal plane, and channel gain. As is discussed further below,
the host controller can fully re-assemble the data stream back into
the individual images captured by each focal plane. In several
embodiments, the imager array includes sufficient storage to buffer
at least a complete row of image data from all focal planes so as
to support reordering and or retiming of the image data from all
focal planes such that the data is always packaged with the same
timing/ordering arrangement regardless of operating parameters such
as (but not limited to) integration time and relative read-out
positions. In a number of embodiments, the imager array includes
sufficient storage to buffer at least a complete line of image data
from all focal planes so as to support reordering and or retiming
of the image data from all focal planes such that the data is
packaged in a convenient manner to ease the host's reconstruction
of the image data, for example retiming/reordering the image data
to align the data from all focal planes to a uniform row start
position for all focal planes irrespective of relative read-out
position.
3.4.2. ASP Sharing
[0145] The imager array illustrated in FIG. 3 includes a separate
ASP associated with each focal plane. An imager array can be
constructed in accordance with embodiments of the invention in
which ASPs or portions of the ASPs such as (but not limited to) the
AFE or the ADC are shared between focal planes. An imager array
that shares ASPs between multiple focal planes in accordance with
embodiments of the invention is illustrated in FIG. 4. The imager
array 300' utilizes an ASP 310' for sampling of all the pixels in
one column of the M.times.N array of focal planes. In the
illustrated embodiment, there are M groups of analog pixel signal
read-out lines connected to M ASPs. Each of the M groups of analog
pixel signal read-out lines has N individual lines. Each of the M
ASPs sequentially processes each pixel signal on its N inputs. In
such a configuration the ASP performs at least N processes per
pixel signal period of the N inputs given that each focal plane at
its input is in an active state. If one or more of an ASP's focal
plane inputs is in an inactive or power down state, the processing
rate could be reduced (so as to achieve a further saving in power
consumption) or maintained (so as to achieve an increase in frame
rate). Alternatively, a common single analog pixel signal read-out
line can be shared by all column circuits in a column of focal
planes (N) such that the time multiplexing function of the ASP
processing can be implemented through sequencing controlled by the
column read-out control block 312'.
[0146] Although the imager array illustrated in FIG. 4 includes
shared ASPs, imager arrays in accordance with many embodiments of
the invention can include dedicates AFEs and share ADCs. In other
embodiments, the sharing ratios of the AFE and ADC do not follow
the same number of focal planes. In several embodiments, each focal
plane may have a dedicated AFE but two or more AFE outputs are
input to a common ADC. In many embodiments, two adjacent focal
planes share the same AFE and one or more of these focal plane
couples would then be input into an ADC. Accordingly, AFEs and ADCs
can be shared between different focal planes in a SOC imager any of
a variety of different ways appropriate to specific applications in
accordance with embodiments of the invention.
[0147] Sharing of ADCs between pairs of focal planes in an imager
array in accordance with embodiments of the invention is
illustrated in FIG. 4d. In the illustrated embodiment, the sharing
of ADCs between pairs of focal planes can be replicated amongst
multiple pairs of focal planes within an imager array. Sharing of
AFEs between pairs of focal planes and sharing of ADCs between
groups of four focal planes in an imager array in accordance with
embodiments of the invention is illustrated in FIG. 4e. The sharing
of AFEs and ADCs illustrated in FIG. 4e can be replicated amongst
multiple groups of four focal planes within an imager array. In
many embodiments, sharing occurs in pairs of focal planes and/or
groups of three or more focal planes.
[0148] In many embodiments, the pixels within each focal plane are
consistently processed through the same circuit elements at all
times such that they have consistent offset and gain
characteristics. In many embodiments, the control and read-out
circuits and AFE are controlled by a common clocking circuit such
that the phases and time slot assignment of each focal plane are
consistent. An example of the phase shift between the column
read-out of the different focal planes in accordance with
embodiments of the invention is illustrated in FIG. 4c. As can be
seen, the read-out of the columns in each focal plane is staggered
to enable processing by a shared ASP in accordance with embodiments
of the invention.
[0149] In order to support a reduction of power when certain focal
planes are not imaging, the ASP, clocking, and bias/current schemes
utilized within the imager array can support multiple sample rate
configurations such that the sampling rate is always P times the
pixel rate of a single focal plane, where P is the number of active
focal planes being processed/sampled.
[0150] A rotated variation of the resource sharing architecture
illustrated in FIG. 4 can also be implemented whereby a single ASP
is shared among all pixels in a row of M.times.N (rather than in a
column of M.times.N). Such an arrangement would, therefore, involve
use of N ASPs each having M inputs or a single input that is common
to the M focal planes, and time-multiplexed by the column read-out
control block using sequencing control.
3.4.3. Column Circuit Sharing
[0151] In another embodiment of the invention, fewer than M*N
column circuits are used for sampling the pixel values of the focal
planes in an imager array. An imager array 301 configured so that
individual focal planes within a column of the imager array share a
common column circuit block 308' such that the device utilizes only
M sets of column circuits in accordance with an embodiment of the
invention is illustrated in FIG. 4a. The M column circuits are
accompanied by M ASPs 310'.
[0152] In several embodiments, the column circuits are time shared
such that they enable read-out of pixels from focal planes above
and below the column circuit. Sharing of a column circuit between
pairs of focal planes within an imager array in accordance with
embodiments of the invention is illustrated in FIG. 4f. The sharing
shown in FIG. 4f is the special case in FIG. 4a, where M=2. Due to
the sharing of the column circuit between the pair of focal planes,
the column circuit operates at twice the rate than the desired
frame rate from a single focal plane. In many embodiments, the
pixels are correlated double sampled and read-out either in their
analog form or analog to digital converted within the column
circuit. Once the last pixel has been shifted out (or the analog to
digital conversion of all the columns has been performed), the
column circuit can be reset to remove residual charge from the
previous pixel array. A second time slot can then be used for the
same operation to occur for the second focal plane. In the
illustrated embodiment, the sharing of ADCs between pairs of focal
planes can be replicated amongst multiple pairs of focal planes
within an imager array.
[0153] In other embodiments, variations on the imager array 301
illustrated in FIG. 4a can utilize more or fewer ASPs. In addition,
the column circuits 308' can be divided or combined to form more or
fewer than M analog outputs for digitization. For example, an
imager array can be designed such that there is a single ASP used
for digitization of the M column circuits. The M outputs of the
column circuits are time multiplexed at the input to the ASP. In
the case that more than M ASPs are used, each of the M column
circuits are further divided such that each column circuit has more
than one analog output for digitization. These approaches offer
trade-offs between silicon area and power consumption since the
greater the number of ASPs, the slower each ASP can be so as to
meet a target read-out rate (frame rate).
[0154] A structural modification to the embodiment illustrated in
FIG. 4a is to split the M column circuits between the top and
bottom of the imager array such that there are M*2 column circuit
blocks. In such a modification each of the M*2 column circuits is
responsible for sampling only half of the pixels of each focal
plane in the column of focal planes (e.g. all even pixels within
each focal plane could connect to the column circuit at the bottom
of the array and all odd pixels could connect to the column circuit
at the top). There are still M*X column sampling, circuits, however
they are physically divided such that there are M*2 sets of X/2
column sampling circuits. An imager array including split column
circuits in accordance with an embodiment of the invention is
illustrated in FIG. 4b. The imager array 301' uses M*2 column
circuit blocks (308a', 308b') and M*2 ASPs (310a', 310b'). As
discussed above, there can also be fewer or more ASPs than the M*2
column circuits. Another variation involving splitting column
circuits in accordance with embodiments of the invention is
illustrated in FIG. 4g in which the column circuit is split into
top/bottom for sampling of odd/even columns and interstitial column
circuits are time shared between the focal planes above and below
the column circuits. In the illustrated embodiment, the splitting
of column circuits and sharing of column circuits between pairs of
focal planes is replicated amongst multiple pairs of focal planes
within an imager array in accordance with embodiments of the
invention. In addition, each of the column circuits can be shared
between an upper and lower focal plane (with the exception of the
column circuits at the periphery of the imager array).
3.4.4. Number and Rate of ASPs
[0155] There are a number of different arrangements for the column
sampling circuitry of imager arrays in accordance with embodiments
of the invention. Often, the arrangement of the ASP circuitry
follows a logical implementation of the column sampling circuits
such that a single ASP is used per column circuit covering X pixels
thus performing X conversions per row period. Alternatively, X ASPs
can be utilized per column circuit performing one conversion per
row period. In a general sense, embodiments of the invention can
use P ASPs per column circuit of X pixels such that there are XIP
conversions per row period. This approach is a means by which the
conversion of the samples in any column circuit can be parallelized
such that the overall ADC conversion process occurs at a slower
rate. For example, in any of the configurations described herein it
would be possible to take a column circuit arrangement that samples
a number of pixels (T) and performs the analog-to-digital
conversion using P ASPs, such that there are T/P conversions per
row period. Given a fixed row period (as is the case with a fixed
frame rate) the individual conversion rate of each ASP is reduced
by the factor P. For example, if there are two ASPs, each runs at
1/2 the rate. If there are four, each ASP has to run at 1/4 the
rate. In this general sense, any number of ASPs running at a rate
appropriate to a specific application irrespective of the
configuration of the column circuitry can be utilized in accordance
with embodiments of the invention.
3.4.5. Row Decoder Optimization
[0156] Imager arrays in accordance with embodiments of the
invention possess the ability to access different rows within each
focal plane at a given instant so as to enable separate operating
parameters with respect to the capture of image information by the
pixels of each focal plane. The row decoder is typically formed
from a first combinational decode of a physical address
(represented as an E bit binary number) to as many as 2.sup.E
"enable" signals (often referred to as a "one-hot" representation).
For example, an 8 bit physical address is decoded into 256 "enable"
signals so as to support addressing into a pixel array having 256
rows of pixels. Each of these "enable" signals are in turn
logically ANDED with pixel timing signals, the results of which are
then applied to the pixel array so as to enable row based pixel
operations such as pixel reset and pixel charge transfer.
[0157] The row decoders can be optimized to reduce silicon area
through sharing of the binary to one-hot decode logic. Rather than
each sub-array having a fully functional row decoder, including
binary to one-hot decoding, many embodiments of the invention have
a single binary to one-hot decoder for a given row of focal planes
within the imager array. The "enable" outputs of this decoder are
routed across all focal planes to each of the (now less functional)
row decoders of each focal plane. Separate sets of pixel level
timing signals would be dedicated to each focal plane (generated by
the row timing and control logic circuitry) and the logical AND
function would remain in each focal plane's row decoder.
[0158] Readout with such a scheme would be performed in time slots
dedicated to each focal plane such that there are M timeslots per
row of focal planes in the camera array. A first row within the
first focal plane would be selected and the dedicated set of pixel
level timing signals would be applied to its row decoder and the
column circuit would sample these pixels. In the next time slot the
physical address would change to point to the desired row in the
next focal plane and another set of dedicated pixel level timing
signals would be applied to its row decoder. Again, the column
circuits would sample these pixels. The process would repeat until
all focal planes within a row of focal planes in the camera array
have been sampled. When the column circuits are available to sample
another row from the imager array, the process can begin again.
3.5. Providing a Memory Structure to Store Image Data
[0159] An additional benefit of the separate control of the capture
of image information by each focal plane in an imager array is the
ability to support slow motion video capture without increasing the
frame rate of the individual focal planes. In slow motion video
each focal plane is read out at a slightly offset point in time. In
a traditional camera, the time delta between frames (i.e. the
capture frame rate) is dictated by the read-out time of a single
frame. In an imager array offering support of independent read-out
time of the individual focal planes, the delta between frames can
be less than the read-out of an individual frame. For example, one
focal plane can begin its frame read-out when another focal plane
is halfway through the read-out of its frame. Therefore an apparent
doubling of the capture rate is achieved without requiring the
focal planes to operate at double speed. However, when outputting
the stream of images from the camera, this overlapping frame
read-out from all focal planes means that there is continuous
imagery to output.
[0160] Camera systems typically employ a period of time between
read-out or display of image data known as the blanking period.
Many systems require this blanking period in order to perform
additional operations. For example, in a CRT the blanking interval
is used to reposition the electron beam from the end of a line or
frame to the beginning of the next line or frame. In an imager
there are typically blanking intervals between lines to allow the
next line of pixels to be addressed and the charge therein sampled
by a sampling circuit. There can also be blanking intervals between
frames to allow a longer integration time than the frame read-out
time.
[0161] For an array camera operating in slow motion capture mode in
accordance with an embodiment of the invention, the frame read-out
is offset in time in all the focal planes such that all focal
planes will enter their blanking intervals at different points in
time. Therefore, there typically will not be a point in time where
there is no image data to transmit. Array cameras in accordance
with embodiments of the invention can include a retiming FIFO
memory in the read-out path of the image data such that an
artificial blanking period can be introduced during transmission.
The retiming FIFO temporarily stores the image data to be
transmitted from all the focal planes during the points in time
where a blanking interval is introduced.
3.6. Imager Array Floor Plan
[0162] Imager arrays in accordance with embodiments of the
invention can include floor plans that are optimized to minimize
silicon area within the bounds of certain design constraints. Such
design constraints include those imposed by the optical system. The
sub-arrays of pixels forming each focal plane can be placed within
the image circle of each individual lens stack of the lens array
positioned above the imager array. Therefore, the manufacturing
process of the lens elements typically imposes a minimum spacing
distance on the imagers (i.e. a minimum pitch between the focal
planes). Another consideration in the focal spacing coming from
optical constraints is the magnitude of stray light that can be
tolerated. In order to limit optical cross-talk between focal
planes, many camera arrays in accordance with embodiments of the
invention optically isolate the individual focal planes from each
other. An opaque barrier can be created between the optical paths
of adjacent focal planes within the lens stack. The opaque barrier
extends down to the sensor cover-glass and can serve the additional
purpose of providing a sensor to optics bonding surface and back
focus spacer. The incursion of the opaque shield into the imaging
circle of the lens can result in some level of reflection back into
the focal plane. In many embodiments, the complex interplay between
the optics and the imager array results in the use of an iterative
process to converge to an appropriate solution balancing the design
constraints of a specific application.
[0163] The space between the focal planes (i.e. the spacing
distance) can be used to implement control circuitry as well as
sampling circuitry including (but not limited to) ASP circuits or
other circuitry utilized during the operation of the imager array.
The logic circuits within the imager array can also be broken up
and implemented within the spacing distance between adjacent focal
planes using automatic place and routing techniques.
[0164] Although specific constraints upon the floor plans of imager
arrays are described above, additional constraints can be placed
upon floor plans that enable the implementation of the various
logic circuits of the imager array in different areas of the device
in accordance with embodiments of the invention. In many
embodiments, requirements such as pixel size/performance, the
optical system of the array camera, the silicon real-estate cost,
and the manufacturing process used to fabricate the imager array
can all drive subtle variations in the imager array overall
architecture and floor plan.
3.6.1. Sampling Diversity
[0165] In many embodiments, the floor plan also accommodates focal
planes that are designed to accommodate an arrangement that yields
a preferred sampling diversity of the scene (i.e. the pixels within
one focal plane are collecting light from a slightly shifted field
of view with respect to other focal planes within the imager
array). This can be achieved through a variety of techniques. In
several embodiments, sampling diversity is achieved by constructing
the imager array so that the focal planes are relatively offset
from the centers of their respective optical paths by different
subpixel amounts through a relative subpixel shift in alignment
between the focal planes and their respective lenses. In many
embodiments, the optical field of view are "aimed" slightly
differently by an angle that corresponds to a subpixel shift in the
image (an amount less than the solid angle corresponding to a
single pixel). In a number of embodiments, slight microlens shifts
between the focal planes is utilized to alter the particular solid
angle of light captured by the microlens (which redirects the light
to the pixel) thus achieving a slight subpixel shift. In certain
embodiments, the focal planes are constructed with pixels having
subtle differences in pixel pitch between focal planes such that
sampling diversity is provided irrespective of optical alignment
tolerances. For example, a 4.times.4 imager array can be
constructed with focal planes having pixels with length and width
dimensions of size 2.0 um, 2.05 um, 2.1 um, 2.15 um and 2.2 um. In
other embodiments, any of a variety of pixel dimensions and/or
techniques for improving sampling diversity amongst the focal
planes within the imager array can be utilized as appropriate to a
specific application.
4. Focal Plane Timing and Control Circuitry
[0166] Referring back to FIG. 1a, imager arrays in accordance with
embodiments of the invention can include focal plane timing and
control circuitry 154 that controls the reset and read-out (hence
integration) of the pixels in each of the focal planes within the
imager array. The ability of an imager array in accordance with
embodiments of the invention to provide flexibility in read-out and
integration time control can enable features including (but not
limited to) high dynamic range imaging, high speed video and
electronic image stabilization.
[0167] Traditional image sensors nominally employ two rolling
address pointers into the pixel array, whose role is to indicate
rows to receive pixel level charge transfer signals as well as "row
select" signals for connecting a given row to the column lines
enabling sampling of the sense node of the pixels. In many SOC
image arrays in accordance with embodiments of the invention these
two rolling address pointers are expanded to 2.times.M.times.N
rolling address pointers. The pointer pairs for each focal plane
can either address the same rows within each focal plane or can be
offset from one another with respect to a global reference.
[0168] Focal plane timing and control address pointer circuitry in
accordance with an embodiment of the invention is illustrated in
FIG. 4h. The focal plane timing and control circuitry 400 includes
a global row counter 402 and read pointer address logic circuitry
404 and reset pointer address logic circuitry 406 associated with
each focal plane. The global row counter 402 is a global reference
for sampling of rows of pixels. In a number of embodiments, the
global row counter 402 counts from 0 to the total number of rows
within a focal plane. In other embodiments, alternative global row
counters are utilized as appropriate to the requirements of a
specific application. The read pointer address logic circuitry 404
and the reset pointer address logic circuitry 406 translates the
global row counter value to a physical address within the array as
a function of settings such as read-out direction and windowing. In
the illustrated embodiment, there are M.times.N read pointer and
reset pointer address logic circuits. Row based timing shifts of
each focal plane read-out and reset positions (FP_offset[x,y]) are
provided to the read pointer address logic and reset pointer
address logic circuits. These timing shifts can be stored in
configuration registers within the imager array. The value of the
timing shifts can be added to the global row counter value (modulo
the total number of rows) before translation to physical addresses
by the read pointer address logic and the reset pointer address
logic circuits. In this way, each focal plane can be provided with
a programmable timing offset. In several embodiments, the timing
offsets are configured based upon different operational modes of
the array camera.
5. System Power Management and Bias Generation
[0169] The system power management bias generation circuitry is
configured to provide current and or voltage references to analog
circuitry such as (but not limited to) the reference voltages
against which an ADC would measure the signal to be converted
against. In addition, system power management and bias generation
circuitry in accordance with many embodiments of the invention can
turn off the current/voltage references to certain circuits when
they are not in use for power saving reasons. Additional power
management techniques that can be implemented using power
management circuitry in accordance with embodiments of the
invention are discussed below.
5.1. Power Optimization
[0170] The master control block of an imager array in accordance
with embodiments of the invention can manage the power consumption
of the imager array. In many embodiments, the master control block
reduces power consumption by "turning off" certain focal planes
during modes of operation where the desired output resolution is
less than the full performance of the device. In such modes,
amplifiers, bias generators, ADCs and other clocked circuits
associated with the focal planes that are not used are placed in a
lower power state to minimize or eliminate static and dynamic power
draw.
5.1.1. Preventing Carrier Migration During Imager Power Down
[0171] Despite a focal plane being in a powered down state, light
is incident upon the pixels in its sub-array. Incident photons will
continue to create charge carriers in the silicon substrate. If the
pixels in a powered-down focal plane are left floating, the charge
carriers will fill the pixel well and deplete the potential barrier
making it unable to trap any further carriers. Excess carriers,
created by the persistent photon flux will then be left to wander
the substrate. If these excess carriers wander from an inactive
focal plane into an active focal plane, and collect in the well of
a pixel in an active focal plane, they would be erroneously
measured to be photo-electrons that were generated within that
pixel. The result can be the appearance of blooming around the
periphery of the active imager caused by the tide of free carriers
migrating into the active focal plane from the inactive
neighbors.
[0172] To mitigate the migration of excess carriers from inactive
focal planes, the photodiodes in the pixels of an inactive focal
planes are connected to the power supply via transistor switches
within each pixel such that the pixel well is held open to its
maximum electrical potential. Holding the well open enables the
photodiode to constantly collect carriers generated by the incident
light and thus reduce the problem of carrier migration from an
inactive imager. The transistors in each pixel are part of the
normal pixel architecture i.e. the transfer gate, and it is the
master control logic along with the row controllers that signal the
transistors to hold the wells open.
5.1.2. Standby Mode
[0173] In many embodiments, reference pixels are used in the
calibration of dark current and FPN. In several embodiments, the
power management circuitry is configured to enable the powering
down of the pixels in a focal plane in such a way that the
reference pixels remain active. In several embodiments, this is
achieved by powering the ASP during the readout of reference pixels
but otherwise maintaining the ASP in a low power mode. In this way,
the focal plane can be more rapidly activated by reducing the need
to calibrate dark current and FPN when the focal plane is woken up.
In many instances, calibration is performed with respect to dark
current and FPN when the reference pixels are powered down during
the low power state of the focal plane. In other embodiments, any
of a variety of partial powering of circuitry can be utilized to
reduce the current drawn by a focal plane and its associated
peripheral circuitry in accordance with embodiments of the
invention.
6. Focal Plane Data Collation and Framing Logic
[0174] Referring again to FIG. 1a, imager arrays in accordance with
several embodiments of the invention include focal plane data
collation and framing logic circuitry that is responsible for
capturing the data from the focal planes and packaging the data
into a container in accordance with a predetermined container
format. In a number of embodiments, the circuitry also prepares the
data for transmission by performing data transformations including
but not limited to any bit reduction to the data (e.g. 10 bit to 8
bit conversion).
[0175] Although specific imager array architectures are described
above, alternative imager array architectures can be used to
implement. Imager arrays based upon requirements, including but not
limited to, pixel size/performance, the optical system of the array
camera, the silicon real-estate cost, and the manufacturing process
used to fabricate the imager array in accordance with embodiments
of the invention. In addition, imager arrays in accordance with
embodiments of the invention can be implemented using any of a
variety of shapes of pixels including but not limited to square
pixels, rectangular pixels, hexagonal pixels, and a variety of
pixel shapes. Accordingly, the scope of the invention should be
determined not by the embodiments illustrated, but by the appended
claims and their equivalents.
* * * * *