U.S. patent application number 12/297581 was filed with the patent office on 2009-12-17 for bandwidth improvement for 3d display.
This patent application is currently assigned to SETRED AS. Invention is credited to Thomas Ericson, Christian Moller, Doug Patterson.
Application Number | 20090309887 12/297581 |
Document ID | / |
Family ID | 38217306 |
Filed Date | 2009-12-17 |
United States Patent
Application |
20090309887 |
Kind Code |
A1 |
Moller; Christian ; et
al. |
December 17, 2009 |
BANDWIDTH IMPROVEMENT FOR 3D DISPLAY
Abstract
A method of operating an autostereoscopic display, the
autostereoscopic display device comprising a switchable aperture
array and a screen, the method comprising: displaying a portion of
an image on the screen for a first period of time; and using the
switchable aperture array to restrict to a second period of time
the time for which a portion of the image is wholly or partly
visible; wherein the second period of time is less than the first
period of time in order to increase the bandwidth of the
autostereoscopic display.
Inventors: |
Moller; Christian; (Oslo,
NO) ; Patterson; Doug; (Kent, GB) ; Ericson;
Thomas; (Bromma, SE) |
Correspondence
Address: |
PATTERSON, THUENTE, SKAAR & CHRISTENSEN, P.A.
4800 IDS CENTER, 80 SOUTH 8TH STREET
MINNEAPOLIS
MN
55402-2100
US
|
Assignee: |
SETRED AS
Oslow
NO
|
Family ID: |
38217306 |
Appl. No.: |
12/297581 |
Filed: |
April 19, 2007 |
PCT Filed: |
April 19, 2007 |
PCT NO: |
PCT/GB2007/001406 |
371 Date: |
January 30, 2009 |
Current U.S.
Class: |
345/522 ;
345/213 |
Current CPC
Class: |
H04N 13/302 20180501;
H04N 13/305 20180501; H04N 13/398 20180501; H04N 13/315
20180501 |
Class at
Publication: |
345/522 ;
345/213 |
International
Class: |
G06F 3/038 20060101
G06F003/038; G06T 1/00 20060101 G06T001/00 |
Foreign Application Data
Date |
Code |
Application Number |
Apr 19, 2006 |
GB |
0607726.7 |
Apr 19, 2006 |
GB |
0607727.5 |
Claims
1.-46. (canceled)
47. A method of operating an autostereoscopic display, the
autostereoscopic display device comprising a switchable aperture
array and a screen, the method comprising: displaying a first
portion of an image on the screen for a first period of time; and
using the switchable aperture array to restrict to a second period
of time the time for which a second portion of the image is wholly
or partly visible; wherein the second period of time is less than
the first period of time in order to increase the bandwidth of the
autostereoscopic display.
48. The method as claimed in claim 47, further comprising
determining the bandwidth by: the spatial resolution of the screen;
the bit depth of each pixel displayed by the screen; the number of
subframes comprising a 3D frame; and the repetition rate of each
subframe.
49. The method as claimed in claim 47, wherein the first period of
time is a minimum time period for display of a pixel on the
screen.
50. The method as claimed in claim 47, further comprising time
multiplexing the screen using a light source of constant
intensity.
51. The method as claimed in claim 47, wherein the second portion
of the image comprises a group of pixels and the image elements of
each pixel and further comprising arranging the image elements to
be displayed in the same order for all pixels in the group such
that the aperture restricts the same image elements for all pixels
in the group.
52. The method as claimed in claim 51, further comprising arranging
the image elements of each pixel in the group to be displayed in
the same order.
53. The method as claimed in claim 51, further comprising arranging
the image elements of each pixel in the group to be displayed with
the same timing.
54. The method as claimed in claim 51, wherein the image elements
are bits.
55. The method as claimed in claim 47, further comprising varying
the length of time by which the switchable aperture array reduces
the period of time that an image shown on the screen is visible to
a viewer.
56. The method as claimed in claim 55, further comprising varying
the length of time by which the switchable aperture array reduces
the period of time that an image shown on the screen in discrete
amounts to define greyscale levels of image brightness.
57. The method as claimed in claim 47, wherein the screen has a
minimum display time for a least significant bit (LSB) and the
switchable aperture array reduces the amount of time that the LSB
is visible.
58. The method as claimed in claim 47, wherein the screen has a
minimum display time for a least significant bit (LSB), and the
switchable aperture array reduces the visible average intensity of
the LSB.
59. A method of operating a time multiplexed autostereoscopic
display device, the autostereoscopic display device comprising a
switchable aperture array and a screen, the screen having variable
output brightness, the method comprising: displaying first
brightness portions of a frame when a particular set of apertures
is open and when the screen is at a first brightness; and
displaying second brightness portions of the frame when the
particular set of apertures is open and when the screen is at a
second brightness.
60. The method as claimed in claim 59, further comprising
displaying a frame as a plurality of sequential subframes, the
first brightness portions comprising a first set of subframes and
the second brightness portions comprising a second set of
subframes.
61. The method as claimed in claim 59, further comprising
displaying one or more sets of additional brightness portions of a
frame when the screen is at one or more additional
brightnesses.
62. The method as claimed in claim 59, further comprising
displaying the first brightness portions and the second brightness
portions of the frame in non-adjacent periods of time.
63. The method as claimed in claim 59, further comprising
synchronizing the switchable aperture array such that a set of
apertures is closed between the times when the first brightness
portions of the frame and the second brightness portions of the
frame are displayed.
64. The method as claimed in claim 59, further comprising
displaying the first brightness portions of the frame
consecutively.
65. The method as claimed in claim 59, further comprising
displaying the second brightness portions of the frame
consecutively.
66. The method as claimed in claim 59, wherein the first brightness
portions of a frame are the most significant bits (MSBs) of an
image.
67. The method as claimed in claim 59, wherein the second
brightness portions of a frame are the least significant bits
(LSBs) of an image.
68. The method as claimed in claim 59, further comprising reducing
the brightness of the screen by reducing the power input into a
light source.
69. The method as claimed in claim 59, further comprising reducing
the brightness of the screen by applying a filter between the light
source and the screen.
70. The method as claimed in claim 59, further comprising arranging
the screen to display different colours sequentially.
71. The method as claimed in claim 59, further comprising applying
colour filters between the light source and the screen to allow
different colours to be displayed on the screen.
72. The method as claimed in claim 71, wherein the colour filters
are part of a colour wheel.
73. The method as claimed in claim 72, further comprising applying
intensity filters in conjunction with colour filters to
sequentially display bright portions and less bright portions of
each colour component of an image.
74. The method as claimed in claim 73, using the intensity filters
to display the first brightness portions of an image and the second
brightness portions of an image consecutively.
75. The method as claimed in claim 59, further comprising
displaying the brightness portions having a lower brightness at a
lower refresh frequency than the brightness portions having a
higher brightness.
76. The method as claimed in claim 59, wherein an image is
displayed and the image is a three dimensional image
77. A method of operating an autostereoscopic display apparatus,
the apparatus comprising a switchable aperture array and a screen,
wherein: a first aperture of the switchable aperture array is open
for a first time period; a second aperture of the switchable
aperture array is open for a second time period; a shared time
space is a period of time between the first and second time
periods; the method comprising: closing the first aperture during
the shared time space; and opening the second aperture during the
shared time space.
78. The method as claimed in claim 77, further comprising closing
the first aperture and opening the second aperture at substantially
the same time and at either the beginning of, the end of, or during
the shared time space.
79. The method as claimed in claim 77, further comprising closing
the first aperture during the shared time space to reduce the
amount of time that an LSB is visible.
80. The method as claimed in claim 77, farther comprising closing
the first aperture at the end of the shared time space and opening
the second aperture at the beginning of the shared time space.
81. The method as claimed in claim 77, further comprising
displaying a first portion of an image for a first time period and
displaying a second portion of the image for a second time period,
such that the first and second portions of the image are displayed
consecutively.
82. The method as claimed in claim 77, wherein the first and second
portions of an image share the shared time space for display of
lowest order bits of each image.
83. The method as claimed in claim 77, wherein the shared time
space is used alternately between the first and second
apertures.
84. The method as claimed in claim 78, wherein the shared time
space is used alternately between the first and second
apertures.
85. The method as claimed in claim 79, wherein the shared time
space is used alternately between the first and second
apertures.
86. A method of operating an autostereoscopic display apparatus,
the apparatus comprising a switchable aperture array and a screen,
the method comprising: displaying a 3D image as a plurality of
consecutively displayed subframes, each subframe displayed on the
screen when a particular set of apertures is open; and displaying a
pixel of an image as the average value of two subframes displayed
for a particular set of apertures.
87. The method as claimed in claim 86, wherein an average value of
a pixel is a value between any two adjacent brightness levels that
are displayed by the autostereoscopic display apparatus in a single
subframe.
88. The method as claimed in claim 86, wherein an average value of
a pixel is an average of a high value and a low value and wherein a
first subframe for the particular set of apertures comprises a
pattern of high values and low values for a plurality of pixels,
and a second subframe for the particular set of apertures comprises
an inverted pattern of the first subframe.
89. A method of operating an autostereoscopic display apparatus,
the apparatus comprising a screen and a switchable aperture array,
the method comprising displaying a three dimensional image by
showing: a first set of images rendered for slits of a first width
and a second set of images rendered for slits of a second width,
wherein a slit comprises one or more adjacent open apertures.
90. The method as claimed in claim 89, further comprising showing
the first set of images with opened slits of the first width and
showing the second set of images with opened slits of the second
width.
91. The method as claimed in claim 47, further comprising altering
the separation between the screen and the aperture array to change
the characteristics of the display apparatus for different
purposes.
92. The method as claimed in claim 59, further comprising altering
the separation between the screen and the aperture array to change
the characteristics of the display apparatus for different
purposes.
93. The method as claimed in claim 77, further comprising altering
the separation between the screen and the aperture array to change
the characteristics of the display apparatus for different
purposes.
94. The method as claimed in claim 86, further comprising altering
the separation between the screen and the aperture array to change
the characteristics of the display apparatus for different
purposes.
95. The method as claimed in claim 89, further comprising altering
the separation between the screen and the aperture array to change
the characteristics of the display apparatus for different
purposes.
96. An autostereoscopic display apparatus arranged to use a method
according to claim 47.
97. An autostereoscopic display apparatus arranged to use a method
according to claim 59.
98. An autostereoscopic display apparatus arranged to use a method
according to claim 77.
99. An autostereoscopic display apparatus arranged to use a method
according to claim 86.
100. An autostereoscopic display apparatus arranged to use a method
according to claim 89.
101. An autostereoscopic display apparatus comprising a central
configuration unit arranged to set, during operation of the
apparatus, at least one of the following: the bit depth of a
displayed image; the range of viewing angles for which viewer
experiences continuous parallax; the apparent depth of the 3D
image; the spatial resolution of the displayed image; the flicker
rate of the displayed image; and the animation rate of the
displayed image.
102. An autostereoscopic display apparatus comprising: a switchable
aperture array wherein during operation a slit width of a parallax
barrier is determined by a number of adjacent apertures opened at
the same time; a screen comprising a two dimensional image source,
the image source capable of displaying a variable frame rate,
and/or a variable pixel bit depth; and an adaptive rendering
apparatus arranged to render images for display on the
autostereoscopic display apparatus according to the configuration
of the autostereoscopic display apparatus.
Description
[0001] The present invention relates to an autostereoscopic display
apparatus. The present invention also relates to a method of
operating an autostereoscopic display.
BACKGROUND
[0002] A well proven method for creating a 3D image is to cause a
viewer to see different perspective views of a scene with each eye.
One way to do this is to display two differently polarized images
on a screen, and for the viewer to wear corresponding polarizing
filters on each eye.
[0003] An autostereoscopic display or a three dimensional (3D)
display may be implemented using an aperture or slit array in
conjunction with a two dimensional (2D) display to display a 3D
image. The principle of the device is that when looking at a 2D
image through a slit array, the slit array separated from the
screen by a distance, then the viewer sees a different part of the
2D image with each eye. If an appropriate image is rendered and
displayed on the 2D display, then a different perspective image can
be displayed to each eye of the viewer without necessitating them
to wear filters over each eye.
[0004] One important parameter which governs quality in most 3D
display technology, is bandwidth, defined as the amount of data
presented by a 3D display. To achieve large depth with high
resolution over a wide viewing area, a large bandwidth is usually
required.
[0005] Embodiments of the invention demonstrate ways in which
bandwidth limitations of autostereoscopic display apparatus may be
overcome in order that high resolution 3D images may be
displayed.
[0006] The invention disclosed herein may be implemented in the
scanning slit time-multiplexed system described in PCT application
PCT/IB2005/001480. However, the invention may also be used in
conjunction with other display systems.
[0007] The scanning slit system creates the 3D effect by showing
different pictures to different locations in front of the display
at high speed. It achieves this by combining a high frame rate 2D
display with a shutter. The shutter is synchronised with the
display and ensures that different portions of the 2D display are
visible only from specific locations. The left image in FIG. 1
shows how a viewer looking through a narrow slit will see two
distinct regions, one for each eye. To create a 3D display from
this simple slit system, the slit must shift laterally sufficiently
quickly so that a viewer sees the scanning shutter as a transparent
window. If all the slits are updated quickly enough to be perceived
as flicker-free, a viewer will see the full resolution of the
underlying 2D display from any position. The 2D display shows
different images synchronised with the opening of slits in the
shutter, as shown in the right image in FIG. 1.
SUMMARY
[0008] Embodiments of the invention are directed towards the field
of improving the bandwidth of an autostereoscopic display.
Bandwidth may be considered as the amount of image information that
can be displayed by the autostereoscopic display over a defined
period of time. An autostereoscopic display may be used to display
animated 3D images, or 3D video. The 3D animation may be computer
generated, in this way perspective views for each frame of the
animation may be readily rendered from basic 3D data associated
with the animated scene.
[0009] Smooth animation is perceived by a viewer if there are at
least 24 frames per second. However, if the screen is refreshed at
this rate, then the viewer will perceive flicker. This is overcome
by refreshing the image displayed on the screen at a higher screen
refresh rate than the animation rate. For example, cinema
projection shows each animation frame twice, resulting in a screen
refresh rate of 48 times per second.
[0010] An autostereoscopic display apparatus uses a switchable
aperture array or shutter array. The switchable aperture array is
an array of switchable slits. The switchable apertures may be
electro-optical and may use Liquid Crystals. In principle, a first
switchable aperture of the array is opened and a correctly rendered
image is displayed behind it. The viewer thus sees different parts
of the image with each eye, each part being a portion of a
different perspective view. The first switchable aperture is closed
and then a second switchable aperture is opened and the process
repeats. In practice, more than one aperture is opened at a time. A
plurality of apertures, each spatially separated from the other is
opened at the same time, and an appropriate image portion displayed
on the screen area behind each. The 2D image displayed on the
screen while an aperture or a group of apertures is open is a
subframe. The minimum number of groups of apertures is determined
by the desired 3D image quality. The number of groups of apertures
determines the number of subframes that must be displayed during a
display refresh time.
[0011] Continuing with the example from cinema projection, where
the display refresh time is 1/48.sup.th of a second. If there are 8
groups of apertures, then 8 subframes are displayed per refresh
frame. This requires a subframe display time of 1/384.sup.th of a
second, or about 2.6 ms.
[0012] A time multiplexed display, such as a Digital Micromirror
Device (DMD), can be used in the 2D display. A DMD typically uses a
fixed intensity light source, and controls the amount of time that
each pixel in a frame is illuminated. This period of time is
interpreted by the viewer as a brightness, the longer the pixel is
illuminated the brighter the pixel is perceived to be. A time
multiplexed display has a minimum period of time that a pixel may
be illuminated on a screen. This provides a limit as to the bit
depth of the image that may be displayed on the screen and in turn
on the autostereoscopic display.
[0013] According to an aspect of the present invention, there is
provided a method of operating an autostereoscopic display, the
autostereoscopic display device comprising a switchable aperture
array and a screen, the method comprising: displaying a portion of
an image on the screen for a first period of time; and using the
switchable aperture array to restrict to a second period of time
the time for which a portion of the image is wholly or partly
visible; wherein the second period of time is less than the first
period of time.
[0014] The first period of time may be a minimum time period for
display of a pixel on the screen. The screen may be time
multiplexed using a light source of constant intensity. The screen
may be time multiplexed and display pixels of constant
intensity.
[0015] There will be a group of pixels for which the same aperture
restricts the image that is visible. For a time multiplexed screen
the image elements (bits) may be arranged in the same order for all
pixels in the group such that the aperture restricts the same image
elements for all the pixels.
[0016] A particular aperture will restrict the time that an area of
the screen is visible. The area of the screen comprises a
particular set of pixels. For a time multiplexed screen the time
components (or bits) of each pixel may be arranged in the same
order of magnitude for all pixels in the particular set of pixels
such that the aperture performs the desired amount of restriction
for all pixels of the particular set of pixels. Further, for a time
multiplexed screen each pixel of the particular set of pixels must
be coordinated such that when the aperture closes, it clips all
pixels at the appropriate time.
[0017] According to an aspect of the present invention, there is
provided a method of operating an autostereoscopic display, the
autostereoscopic display device comprising a switchable aperture
array and a screen, the method comprising: using the switchable
aperture array to restrict a period of time that an image shown on
the screen is visible to a viewer.
[0018] According to an aspect of the present invention, there is
provided a method of operating an autostereoscopic display, the
autostereoscopic display device comprising a switchable aperture
array and a screen, the method comprising: using the switchable
aperture array to reduce the intensity of the image visible to a
viewer.
[0019] The extent to which the switchable aperture array reduces
the period of time that an image shown on the screen is visible to
a viewer may be varied. The length of time by which the switchable
aperture array reduces the period of time that an image shown on
the screen is visible to a viewer may be varied. This length of
time may be varied in discrete amounts to define greyscale levels
of image brightness.
[0020] According to an aspect of the present invention, there is
provided a method of operating an autostereoscopic display, the
autostereoscopic display device comprising a switchable aperture
array and a screen, wherein the screen has a minimum image display
time, the method comprising: using the switchable aperture array to
reduce the amount of time that an image displayed on the screen is
visible below the minimum image display time.
[0021] According to an aspect of the present invention, there is
provided a method of operating an autostereoscopic display, the
autostereoscopic display device comprising a switchable aperture
array and a screen, the method comprising: displaying a particular
frame of a scene on the screen for a first period of time; and
using the switchable aperture array to allow a portion of the
screen to be visible to a viewer for a second period of time;
wherein: the second period of time begins before the first period
of time; or the second period of time ends after the first period
of time; such that for a portion of the second period of time a
frame either immediately preceding or immediately following the
particular frame is visible on the portion of the screen.
[0022] According to an aspect of the present invention, there is
provided a method of operating a time multiplexed autostereoscopic
display, the autostereoscopic display device comprising a
switchable aperture array and a screen, the screen having variable
output brightness, the method comprising: displaying bright
portions of a frame when the screen is at a full brightness and
then displaying less bright portions of the frame when the screen
is at a reduced brightness.
[0023] The bright portions of the frame and the less bright
portions of the frame may be displayed in non-adjacent periods of
time. The switchable aperture array may be synchronised such that a
set of apertures is open when the bright portions of the frame and
the less bright portions of the frame are displayed. The switchable
aperture array may be synchronised such that a set of apertures is
closed between the times when the bright portions of the frame and
the less bright portions of the frame are displayed. The bright
portions of all subframes of a three dimensional image may be
displayed adjacent in time. The less bright portions of all
subframes of a three dimensional image may be displayed adjacent in
time.
[0024] The bright portions of a frame may be the most significant
bits (MSBs) of an image. The less bright portions of a frame may be
the least significant bits (LSBs) of an image. There may be more
than one level of bright portions and more than one level of less
bright portions that may all be displayed with different levels of
brightness.
[0025] According to an aspect of the present invention, there is
provided a method of operating a time multiplexed autostereoscopic
display, the autostereoscopic display device comprising a
switchable aperture array and a screen, the screen having variable
output brightness, the method comprising: displaying first
brightness portions of a frame when the screen is at a first
brightness and then displaying second brightness portions of the
frame when the screen is at a second brightness. The method may
further comprise displaying one or more sets of additional
brightness portions of a frame when the screen is at one or more
additional brightnesses. The first, second and additional
brightness levels may be different.
[0026] The brightness of the screen may be reduced by reducing the
power input into a light source. The brightness of the screen may
be reduced by applying a filter between the light source and the
screen. The screen may be arranged to display different colours
sequentially. Colour filters may be applied between the light
source and the screen to allow different colours to be displayed on
the screen. The colour filters may take the form of a colour wheel.
Intensity filters may be used in conjunction with colour filters to
sequentially display bright portions and less bright portions of
each colour component of an image.
[0027] The screen may display different colour components of an
image concurrently. Intensity filters may be used to display the
bright portions of an image and the less bright portions of an
image consecutively.
[0028] According to an aspect of the present invention, there is
provided a method of operating an autostereoscopic display, the
autostereoscopic display device comprising a switchable aperture
array and a screen, the method comprising: splitting a frame into a
plurality of subframes. Each subframe represents a different
portion of the frame. Each subframe may be different. The
switchable aperture array is synchronised such that a plurality of
apertures are open for each subframe. The subframes are shown in
succession at a fast rate such that a viewer perceives the sum of
the plurality of subframes to be the same image as the original
frame. The viewer perceives the sum of the plurality of subframes
due to persistence of vision, if the rate of display of sequential
subframes is sufficiently fast.
[0029] More than one subframe may be displayed for a particular
group of opened apertures. A first subframe contains the LSBs and a
second subframe does not contain the LSBs. Alternatively, a first
selection of pixels in the first subframe may contain the LSBs and
a second selection of pixels in the second subframe may contain the
LSBs, the second selection of pixels being the inverse selection of
the first selection of pixels. The first selection of pixels may
comprise every other pixel of the screen, in a chess board pattern.
The pixel selection may be a high frequency pattern where one
subframe contains the pattern and one subframe contains the inverse
of the pattern.
[0030] A first aperture is closed and a second aperture is opened
at substantially the same time, this time is the switching time.
The switching time may be at the start of, or end of, or during, a
shared time space. The shared time space is a time period between
the first and second time periods.
[0031] The switchable aperture array may switch between a
transparent state and an opaque state during a shared time space.
The area of screen displaying a first portion of an image for a
first time period is used to display a second portion of an image
for a second time period. The shared time space is a time period
between the first and second time periods. A first aperture is
closed and a second aperture is opened at substantially the same
time, this time is the switching time. The switching time may be at
the beginning, during, or at the end of the shared time.
[0032] The first and second portions of an image are adjacent in
time. Accordingly, the first and second portions of an image share
the same time space for display of the lowest order bits of each
image. Alternatively, the shared time space is used alternately
between the first and second shutters.
[0033] According to an aspect of the present invention, there is
provided a method of operating an autostereoscopic display
apparatus comprising a first and second projector, each projector
using light of a different polarization, a screen which maintains
the polarization of light, a first polarizing shutter and a second
polarizing shutter, the method comprising selectively switching the
polarization state of the first and second polarizing shutters to
selectively display an image from one projector on a particular
portion of the screen to a viewer.
[0034] According to an aspect of the present invention, there is
provided an autostereoscopic display apparatus comprising:
[0035] a first and second projector, each projector using light of
a different polarization,
[0036] a screen which maintains the polarization of light,
[0037] a first polarizing shutter and
[0038] a second polarizing shutter,
[0039] wherein the polarization state of the first and second
polarizing shutters is selectively switched to selectively display
an image from one projector on a particular portion of the screen
to a viewer.
[0040] According to an aspect of the present invention, there is
provided an autostereoscopic display apparatus comprising:
[0041] a screen;
[0042] a first projector arranged to operate with light polarized
in a horizontal direction;
[0043] a second projector arrange to operate with light in a
vertical direction;
[0044] a first switchable polarization array arranged to
selectively rotate the polarization of light passing therethrough;
and a second switchable polarization array arranged to selectively
rotate the polarization of light passing therethrough.
[0045] According to an aspect of the present invention, there is
provided an autostereoscopic display device comprising a screen and
a switchable aperture array, the screen displaying a plurality of
images concurrently, each image comprising a different light
bundle, and each aperture of the switchable aperture array
cooperating with an interference filter. Each interference filter
may be arranged to pass the light of one light bundle. Each light
bundle may be a set of distinct red, green and blue light
frequencies.
[0046] According to an aspect of the present invention, there is
provided an autostereoscopic display apparatus comprising:
[0047] a plurality of 2D image generators, each image generator
using light of a different characteristic,
[0048] a screen which maintains the characteristic of light from
each 2D image generator,
[0049] a switchable aperture array, each aperture comprising a
filter
[0050] wherein the apertures are selectively switched to
selectively display an image from a 2D image generator on
particular portion of the screen to a viewer.
[0051] Each 2D image generator may be a projector. The
characteristic of light may be a polarization. The characteristic
if light may be a frequency. The characteristic of light may be a
light bundle.
[0052] Each aperture of the aperture array may have an associated
lens. The lens may be placed on the same side of the shutter as the
screen, or on the opposite side of the shutter to the screen. Each
aperture of the aperture array may have two associated lenses, one
on each side of the aperture.
[0053] Each aperture of the aperture array may have an associated
holographic element. The holographic element may be placed on the
same side of the shutter as the screen, or on the opposite side of
the shutter to the screen. Each aperture of the aperture array may
have two associated holographic elements, one on each side of the
aperture.
[0054] The screen may comprise an asymmetric optical diffuser. A
plurality of images may be projected onto the screen with different
angles of incidence such that a different image is viewed on the
diffuser dependent on the angle of observation of the diffuser.
Different angles of incidence may be achieved using a plurality of
projectors. Different angles of incidence may be achieved from a
single projector using at least one mirror to create a plurality of
optical paths between the projector and the diffuser.
[0055] Head tracking apparatus may be used to monitor the position
of a viewer, the image displayed by the autostereoscopic display
apparatus is then rendered according to the detected position of
the user.
[0056] The screen may comprise two diffusive elements, a first
diffusive element and a second diffusive element, the first
diffusive element arranged between the second diffusive element and
the aperture array. The first diffusive element is transparent to
light from the second diffusive element. The second diffusive
element displays background images to provide an increased depth of
field for the autostereoscopic display.
[0057] The aperture array may comprise black stripes between
scanned apertures. For a given number of scanned apertures, black
stripes introduced between them results in narrower apertures. The
black stripes may be implemented by closing a first set of
apertures and only scanning a second set of apertures of a
switchable aperture array. This results in improved depth
resolution.
[0058] The aperture array may comprise average value apertures
between scanned apertures. For a given number of scanned apertures,
average value apertures introduced between them results in narrower
apertures. The average value apertures may be implemented by
opening an average value aperture before the end of the period of
time that a first adjacent scanned aperture is open, and closing
the average value aperture during a period of time that a second
adjacent scanned aperture is open. The length of time that the
average value aperture is open may have a mid-point in time that is
coincident with the time that the second adjacent aperture is
opened. The length of time that the average value aperture is open
may have a mid-point in time that is coincident with the time that
the first adjacent aperture is closed.
[0059] The average value apertures may be implemented by opening an
average value aperture half way into the period of time that a
first adjacent scanned aperture is open, and closing the average
value aperture half way into the period of time that a second
adjacent scanned aperture is open. The first and second adjacent
scanned apertures are on opposite sides of the average value
aperture.
[0060] According to an aspect of the present invention, there is
provided a method of operating an autostereoscopic display, the
autostereoscopic display device comprising a switchable aperture
array and a screen, the method comprising:
[0061] displaying a first frame of a scene on the screen for a
first period of time;
[0062] opening a first aperture of the switchable aperture array
for the first period of time;
[0063] displaying a second frame of a scene on the screen for a
second period of time;
[0064] opening a second aperture of the switchable aperture array
for the second period of time;
[0065] opening an intermediate aperture during the first period of
time; and
[0066] closing the intermediate aperture during the second period
of time, wherein the intermediate aperture is between the first and
second apertures.
[0067] The autostereoscopic display apparatus displays a three
dimensional image as a plurality of subframe. Each subframe is
rendered to correspond to at least one open slit in the aperture
array. A subframe may comprise a plurality of strips of rendered
images, each strip rendered for a particular slit. For each
subframe a plurality of spatially separated slits are consecutively
opened and a rendered image strip is displayed on the screen behind
each open slit. A slit may comprise one or more apertures. The more
apertures a slit comprises, the wider the slit. A three dimensional
image may be displayed by showing a first set of subframes having
slits of a first width and a second set of subframes having slits
of a second width.
[0068] According to an aspect of the present invention, there is
provided an autostereoscopic display apparatus comprising a central
configuration unit arranged to set, during operation of the
apparatus, at least one of the following:
[0069] the bit depth of a displayed image;
[0070] the range of viewing angles for which viewer experiences
continuous parallax;
[0071] the apparent depth of the 3D image;
[0072] the spatial resolution of the displayed image;
[0073] the flicker rate of the displayed image; and
[0074] the animation rate of the displayed image.
[0075] According to an aspect of the present invention, there is
provided an autostereoscopic display apparatus comprising:
[0076] a switchable aperture array wherein during operation the
slit width of a parallax barrier is determined by a number of
adjacent apertures opened at the same time;
[0077] a screen comprising a 2D image source, the image source
capable of displaying a variable frame rate, and a variable pixel
bit depth; and
[0078] an adaptive rendering apparatus arranged to render images
for display on the autostereoscopic display apparatus according to
the configuration of the autostereoscopic display apparatus.
[0079] The autostereoscopic display apparatus has a shutter array.
A first and second switchable aperture array may form the shutter
array. The shutter array cooperates with a display screen to create
a display apparatus. An arrangement may be provided to alter the
separation between the display screen and the shutter array to
change the characteristics of the display apparatus for different
purposes. The arrangement may be a simple electromechanical
arrangement comprising motors, worm gears and racks at each corner
of the display apparatus.
BRIEF DESCRIPTION OF THE DRAWINGS
[0080] Non-limiting embodiments of the present invention will be
described by way of example with reference to the accompanying
drawings, in which:
[0081] FIG. 1 illustrates a viewer looking at a screen through a
slit;
[0082] FIG. 2 shows a shared time space between consecutive
subframes;
[0083] FIG. 3 shows shared time space being used for alternate
subframes in consecutive cycles;
[0084] FIG. 4 shows shared time space being equally shared between
subframes in consecutive cycles;
[0085] FIG. 5 shows superimposed horizontal and vertical
polarization systems;
[0086] FIG. 6 shows a shutter in combination with a lenticular;
[0087] FIG. 7 shows a comparison between an traditional directional
diffuser and a directional diffuser;
[0088] FIG. 8 shows a projector arrangement suitable for use with a
directional diffuser;
[0089] FIG. 9 shows a further arrangement suitable for use with a
directional diffuser;
[0090] FIG. 10 shows an arrangement comprising two diffusers;
[0091] FIG. 11 shows a narrow slit arrangement with odd apertures
always closed;
[0092] FIG. 12 shows a the operation of the odd numbered slits as
average value slits;
[0093] FIG. 13 illustrates a pixel on the screen sweeping a
narrower volume of space in the 3D scene, this providing improved
resolution;
[0094] FIG. 14 shows the image cones for a pixel for two adjacent
slits;
[0095] FIG. 15 shows a bit sequence wherein all bits are centred in
time about a mid-point of the subframe duration;
[0096] FIG. 16 shows the operation of a central configuration
unit;
[0097] FIG. 17 shows the viewing region where continuous parallax
is available;
[0098] FIG. 18 shows a shutter arrangement where the slit width
equals the width of two switchable apertures;
[0099] FIG. 19 shows a frame cycle comprising a subframe displayed
for each of 6 groups of slits;
[0100] FIG. 20 shows a frame cycle comprising 2 subframes displayed
for each of 6 groups of slits; and
[0101] FIG. 21 shows a frame cycle comprising 9 subframes displayed
for 9 slit groups having a slit width of 11 and 3 subframes
displayed for 3 slit groups having slit widths of 31.
DETAILED DESCRIPTION OF THE DRAWINGS
[0102] Types of Bandwidth Improvements
[0103] 3D display systems can be flexible in the sense that
bandwidth may be prioritized in different ways depending on
application. The overall bandwidth is defined as the total number
of addressable pixels and the number of colour bits per addressable
pixel. In the time multiplexed system described above bandwidth is
the combination of four factors: [0104] 1. The XY resolution for a
single frame on the underlying display [0105] 2. The colour bit
depth for each pixel [0106] 3. The number of unique frames
presented within a full 3D frame [0107] 4. The repetition rate for
each unique frame
[0108] For a given bandwidth one may choose to prioritise between
points 2 and 3 above. For example, if colour depth is reduced the
number of unique frames can increase giving an image with better
depth or larger viewing region.
[0109] Depending on the 3D data being shown and the audience, one
may want to make different trade offs in how the bandwidth is
prioritised. Hence, a method of enabling changing this trade off
dynamically would be valuable. This may be achieved by adding
dynamic control to several parts of the 3D display system. The user
may then control settings through a software or other input
interface.
[0110] From a user perspective the main properties that may be
changed are: [0111] 1. The colour or greyscale bit depth [0112] 2.
The size or angle of the zone where a viewer experiences continuous
parallax [0113] 3. The depth quality of the image [0114] 4. The XY
(horizontal and vertical) spatial resolution [0115] 5. The flicker
rate [0116] 6. The animation rate
[0117] One way to implement such flexibility is to have a central
configuration unit that sends instructions on system settings to
the different system components. An example of a block diagram
using a control panel as a central configuration unit is shown in
FIG. 16. The unit can be a PC that is running the 3D application
being used. It can send instructions either through a separate
communication channel for changing settings or embedded in existing
synchronisation and data transfer channels. The operation will be
explained by way of example.
[0118] In a base example the display has 5 bits greyscale bit
depth, the angle with continuous parallax for a given slit is 45
degrees and a given depth quality. The setup is shown in FIG. 17.
This is a result of a specific setup where:
[0119] d=distance between the shutter and the underlying display
plane
[0120] l=width of a slit in the shutter
[0121] N=number of unique frames shown by the underlying display.
In this case this also equals the number of slits of slit width l,
between simultaneously open slits. In this example it is N=6.
[0122] The complexity of the example is limited in order to
simplify explanation. N may be significantly larger and the image
portion behind an aperture does not be to be centred behind the
aperture. Edge effects are not described in detail. Also, the
actual angle with continuous parallax experienced by viewers may
not be the same as the angle with continuous parallax for a
particular slit or aperture.
[0123] The user decides to increase the size of the zone with
continuous parallax to around 80 degrees. However, this cannot be
achieved without compromising another property. In a first example,
shown in FIG. 18, the depth quality is reduced while the greyscale
bit depth is maintained: [0124] 1. The central configuration unit
stores the input from the user to increase the viewing zone from
around 45 to around 80 degrees [0125] 2. Via an I2C channel, or
other interface, it sends instructions to the electronics control
unit for the shutter. The shutter driving sequence is changed so
that two or more adjacent columns are switched simultaneously to
give a slit width l, twice the width of what it was in the initial
state. Another way to achieve the same effect is to send
instruction to a mechanism that reduces the distance d, between the
shutter and the underlying display. [0126] 3. The underlying
display, e.g. a projector, receives the new settings through
communication over I2C or other communication channel. It is
instructed to maintain the same number of unique frames and the
same greyscale bit depth. Because the same number of unique frames
are shown over a larger viewing zone the depth quality will be
reduced. The underlying display may require new instructions on the
size and coordinates of an image portion shown behind a slit.
[0127] 4. The rendering engine receives the new setting. It changes
the coordinates of frustums used in rendering image data and the
size of image portions sent to the underlying display. In this
example the size of the image portions doubles. It may also use the
information to change the rendering method or other filters to
optimise the image quality for the specific setting. [0128] 5. The
software application or other content source receives the new
setting. It may for example use this information to include more
data in the scene given the higher maximum viewing angle in a
scene.
[0129] In a second example the viewing zone is increased and
greyscale bit depth is reduced in order to maintain the depth
quality. [0130] 1. The central configuration unit stores the input
from the user to increase the viewing zone from around 45 to around
80 degrees and to reduce greyscale bit depth [0131] 2. The
underlying display or screen, e.g. a projector, is instructed to
change its imaging sequence to reduce bit depth from 5 bits to 4
bits. In this example this allows for the number of unique frames
N, to double from 6 to 12. The underlying display may require new
instructions on the size and coordinates of an image portion shown
behind a slit. [0132] 3. Instructions are sent to the electronics
control unit for the shutter. The shutter driving sequence is
changed so that instead of having 6 groups of slits that are
synchronised with display frames there are 12 groups of slits that
are synchronised with display frames. The time that a slit is open
is reduced to half to be the same duration as a frame in the
underlying display. The slit width l, is kept constant. [0133] 4.
The rendering engine receives the new setting. In response it
changes the frustums used in rendering image data and the size of
image portions sent to the underlying display are doubled. [0134]
5. The software application or other content source receives the
new setting. It may for example use this information to include
more data in the scene given the higher maximum viewing angle in a
scene.
[0135] From the above example one can see that a few elements are
typical for achieving a flexible system: [0136] A shutter and
shutter electronics that can vary effective slit width and sequence
of switching. One way to achieve this is to have many very narrow
slits that can be switched in groups to create wider effective slit
widths. [0137] A flexible image source as underlying display. For
example the image source may allow increasing the frame rate by
reducing colour or greyscale bit depth. The frame rate may also be
increased by reducing flicker rate, animation rate or spatial
resolution. [0138] An adaptive rendering solution that may provide
image data to the image source based on the chosen display
setting.
[0139] Additionally one may add a mechanism for varying the
distance d, between the shutter and the underlying display.
[0140] The user may be given control to change any of the above
properties with small or continuous increments. In some situations,
it may be desirable to offer a number of presets instead. One
example of this could be to have a single user setting and a
multiple user setting where a number of properties are changed when
switching between the two presets.
[0141] Two areas of improving bandwidth will be addressed:
[0142] Increasing display bandwidth--this looks at how the
bandwidth of the basic 3D display setup can be increased
[0143] Increasing system bandwidth--this looks at how the bandwidth
can be increased further and used more efficiently by modifying the
basic principle of the display.
[0144] Increasing Display Bandwidth
[0145] To fully explain the methods for increasing the system
bandwidth in the following section it is useful to give a
background on the additional restrictions and possibilities that
arise from a 3D display compared to a 2D display.
[0146] In a 2D display system, such as a cinema projector, a frame
is typically an image in a time series of images of a scene. The
frame duration is set such that images are updated sufficiently
quickly to give smooth animation. In cinema this animation rate is
24 frames per second. However, if an image or any light source is
updated at only this animation rate the eye typically perceives
flicker. That is why in a cinema projector every frame is shown
twice in succession to give an overall refresh rate that is
sufficiently high not to give flicker.
[0147] In a 3D system each animation frame is made up of a number
of subframes, essentially representing different perspectives of
the scene. In a time multiplexed system these are shown in a rapid
sequence. Hence, the duration of each subframe will be shorter than
the overall frame increasing demands on response time. Furthermore,
the subframes must be repeated and distributed in a way that does
not give rise to any frequency elements that are perceived as
flicker. Typically this is solved by running the sequence of the
subframes at a rate such that the duration of the full 3D frame
exceeds the animation rate. Compared to a 2D display this gives
rise to some significant differences: [0148] The duration of which
a subframe must display the information is shorter than in a 2D
system. This increases demands on response time. In a field
sequential colour system it also forces bit durations to be reduced
in order to maintain the same bit depth. [0149] Each subframe is
repeated at a rate that is typically higher than the minimum
animation rate. Unlike in a 2D system this cannot be solved simply
by showing the same subframe twice in rapid succession, because one
would still have a frequency component at the animation rate. As a
result the same subframe is typically repeated at regular
intervals.
[0150] Reducing Length of Bits
[0151] For a time multiplexed image source, such as a DMD,
bandwidth is partly determined by the shortest possible duration of
the least significant bit (LSB). When a fixed intensity light
source is used, subsequent bits are typically power-of-two
multiples of the LSB duration. Reducing the LSB duration therefore
allows increased bit depth or increased number of frames per second
or both.
[0152] In some instances one may want to increase the bandwidth to
a level where the image source cannot support a sufficiently short
LSB. One way to achieve this is to have the image source and the
optical shutter synchronised with another device that also
modulates the light. There are several options on how to do this:
[0153] 1. Modulate the intensity of the light before the imaging
device [0154] 2. Modulate the length of a light pulse before the
imaging device [0155] 3. Modulate the intensity of the light after
the imaging device [0156] 4. Modulate the length of a light pulse
after the imaging device
[0157] The above methods can be combined. They can be applied to
part of or the entirety of the imaging device. For a self-luminous
image device, methods 3 and 4 can be used. One way to implement 1
or 2 above is to have a light source that is synchronised with the
image source. If, for example all the LSBs on the whole image
device are placed in the same time window, the light source could
be switched off before the end of the LSB, providing a light burst
which is shorter than the LSB that the image source can provide,
and thereby reducing the intensity of the LSB. The light source
could also be dimmed for the duration of the LSB to achieve the
reduced intensity. The light source could for example be an LED or
an LCD backlight. It could also be a combination of a constant
light source and an LED which provides the variation in intensity.
Instead of varying the light source one could have a variable
filter, e.g. an intensity wheel, between the light source and the
imaging device to give the same effect. One could also choose to
have different intensities for other bits or groups of bits. In an
extreme case the light intensity would be unique for each bit
plane.
[0158] One could choose to split a subframe into two or more
partial subframes. For example all the partial subframes with high
order bits could be shown in a group at a higher light intensity
and then the partial subframes with lower order bits could be shown
in a group with lower light intensity. This way the required speed
at which the light source must switch intensity is reduced compared
to a case where each subframe is not split. Clearly the shutter
sequence must change such that the correct slit opens for each
partial subframe.
[0159] A variation of the above principle is to have two light
sources with different intensity levels. A shutter can be used to
switch between the light sources so that they illuminate the
imaging device for alternate frames or part of frames.
[0160] One way to implement 3 and 4 above is to use a shutter or
filter after the imaging device. In a scanning slit system there is
already a shutter in place, which could be used for this purpose.
If the shutter goes from transparent to blocking light such that
the LSB from the imaging device is cut off, the LSB is again
reduced. It could also have a grey state which would reduce the
intensity of the LSB. The above methods are not restricted to the
LSB. It is possible to vary the light intensity for each bit.
[0161] The eye is less sensitive to flicker for low light
intensities. Therefore it is possible to show less significant bits
at lower frequencies than more significant bits. For example, a
certain frame rate might require an LSB of shorter duration than
the image source can provide. Restricting the LSB to every other
frame allows its duration to be doubled, satisfying the image
source's minimum LSB duration requirement. This method is not
restricted to the LSB and could be extended to more significant
bits. The LSB or other bit may be present in fewer than every other
frame, i.e. display of the LSB could skip two or more frames.
[0162] There is more than one way of implementing the above method.
The overall frame duration could be kept constant, such that the
frames containing the LSB are the same length as those that do not
contain the LSB. In the frames that do not contain the LSB, the
time window for the LSB will be replaced with dark time.
Alternatively the overall frame duration could vary between frames
with the LSB and frames without the LSB. This could be supported by
a shutter where each slit can be open for different time periods.
For example, if only every other frame contains the LSB the time
period for the shutter will vary between t for frames without the
LSB and (t+LSB duration) for frames with the LSB.
[0163] The method could be implemented through an overlay of an
alternating spatial pattern. An example of this would be an
alternating checkerboard pattern such that for one frame every
other pixel displays the LSB and every other pixel does not display
the LSB. In the next frame the checkerboard pattern is inverted and
the pixels that in the previous frame displayed the LSB do not
display the LSB and vice versa. Overall, in this example every
pixel will have the LSB present in every other frame. This method
can reduce the overall perception of flicker. Many different
patterns can be used where the LSB is on average present in a
fraction of every frame.
[0164] Increase Grey Scale Bandwidth
[0165] At some point the imaging device will not support shrinking
the LSB further to gain more bandwidth. In some applications it is
desirable to have higher bit depth in grey scale than in colour.
For example, a medical x-ray may contain very high bit depth
greyscale information, while colour bit depth may not be as
important. This can be achieved through a setup that allows
switching between a mode where different optical circuits provide
different base colours and another mode where different optical
circuits provide different white light intensity.
[0166] One way to achieve this is to be able to switch between
colour filters and static intensity filters. The latter could cover
an adjacent but non-overlapping range of intensity values. By way
of example: a 15 bit greyscale range can be achieved using three
5-bit greyscale chips by applying 1/32.times. and 1/1024.times.
intensity filters to two of the chips. Send the top five bits to
the unfiltered chip, the middle five bits to the 1/32x.times. chip,
and the last five bits to the 1/1024.times. chip. An alternative
way of achieving different intensity levels is to use a single
light source and beam splitters. Yet another method is to use
different intensity light sources. An example of this would be
using an LED light source for lower brightness projector. This
would also allow the lower brightness projector to use light
modulation as explained above.
[0167] Of course, you could also just use two optical circuits to
get 10 bits of greyscale, but sticking to three allows the
possibility of mechanically switching filters to give a 15 bit RGB
system.
[0168] An electronic input board can be designed such that it can
split an RGB input signal into either different colour signals or
into different greyscale bands.
[0169] There are several ways of implementing the distribution of
the electronic signals to the imaging devices. One method is having
a central input board, which distributes the data appropriately to
all the available imaging devices and synchronises these. Another
method involves multiple input boards that are synchronised, and
which in turn distribute the data and synchronise the imaging
devices.
[0170] Increase Colour Bandwidth
[0171] In a similar fashion to using more than one chip to achieve
higher greyscale levels, more than 3 optical circuits can be used
to increase the bit depth for each base colour. For example,
another setup would use 6 or more optical circuits to give 24 bit
RGB at 3000 fps, by apportioning 4 bits of the 24 bit value to each
projector.
[0172] Yet another setup could include a colour wheel for one
optical circuit and intensity filters for other optical circuits.
Through this method it is possible to have a higher greyscale bit
depth than full colour bit depth.
[0173] Sharing Time Space Between Frames
[0174] In some instances it is acceptable that two subframes that
are adjacent in time share the same time space for lower order bits
as shown in FIG. 2. For example this could mean that if one
subframe has the LSB set to 0, the next subframe must also have the
LSB set to 0. It could also mean that the subframes alternate the
use of the time space.
[0175] By allowing this one could in some instances achieve more
effective implementations of some of the principles described
above. Implementations of the principle include, but are not
restricted to, the following: [0176] 1. Alternate use of time space
between consecutive subframes. A shutter after the imaging device
alternates switching such that in one cycle c=1 the bit or bits in
the shared time space belongs to one subframe, e.g. subframe 2, and
in the next cycle to the adjacent subframe, e.g. subframe 1. FIG. 3
illustrates how this is used for the shared time space between
frame 1 and frame 2. [0177] 2. Another example is when the shared
time space is the shortest light pulse that the imaging device can
support. The shutters can then be used to reduce the pulse further.
By sharing the time space it is in some instances possible to
increase the subframe display rate. FIG. 4 illustrates how this is
implemented between subframe 1 and subframe 2. [0178] 3. Another
example of shared time space is described below in the section
titled "Effective use of bandwidth".
[0179] The two above implementations can also be combined by using
the shutters to cut off the LSB and then alternating which subframe
shows the LSB+1.
[0180] Increased System Bandwidth
[0181] Filters for Superimposed System
[0182] The methods above involve showing only one image on the
image plane at any one point in time. In order to increase the
bandwidth further one can show multiple images at any one point in
time. A general solution could be comprised of a set of images
superimposed on the image plane. The shutter would then contain
filters which selects only one or a subset of the images for a
particular slit or aperture.
[0183] Polarization
[0184] One example of this is superimposing light with different
polarization. Using two projectors, one with vertical and one with
horizontal polarization in conjunction with a diffuser that
maintains the polarization of the light, one can design the shutter
such that these will act as two independent systems that are
superimposed in the same space. FIG. 5 shows one example of such a
system.
[0185] In this example Shutter A and Shutter B represent liquid
crystal cells. Consider an area in the centre of the display for a
given point in time. In the centre there is a strip from the
horizontally oriented projector (H), which is synchronized with the
opening of slit 7 on Shutter A. The cone from slit 7 must hence be
open for horizontally polarized light only. Slits 5,6,8 and 9
should be closed for any polarization. The cones from slits 4 and
10 on the other hand are open for vertically polarized light only.
This way, the region H is completely overlapped by the two areas V,
which means that two independent images can be projected to give
double system bandwidth.
[0186] The operation is as follows. Shutter B does not twist the
light for slits 6,7 and 8. This means that light from the regions V
but not from H are filtered out for these slits by Polarization
Filter B. Slits 3,4,5 and 9,10,11 on the other hand twist the light
to filter out light from the region H but not from V. All light is
now horizontally polarized. Slits 4,7 and 10 on Shutter A are set
to twist the polarization of the light so that it passes through
the vertical filter at the slit. Slits 5,6,8 and 9 are set not to
twist the polarization so the light is blocked by the vertical
filters.
[0187] Note that Shutter B does not give dark zones, since all
light exits as horizontally polarized. This means that one will see
adjacent regions when going outside the maximum viewing angle. A
third shutter could be added to block this cross-talk if
desired.
[0188] Shutter B could be replaced with a static compensation film.
The film would have stripes twisting the polarization interlaced
with stripes not twisting the polarisation. In this case one could
choose to make the stripes one slit wide and put them as close as
possible to Shutter A.
[0189] With a liquid crystal with symmetric rise and fall time that
can be used in both normally white and normally black mode it would
be possible to have a single shutter in the above system. It would
use alternating polarization filters for each slit.
[0190] Colour Filters
[0191] A similar approach may be used having multiple projectors in
conjunction with complementary RGB light filters. Each projector
projects light of a particular Red, Green and Blue frequency. The
red frequency, green frequency and blue frequency define a light
bundle. Devices for projecting such colour images are known. These
projectors may be combined with interference filters in the
shutter. Display types other than projectors could be used in a
similar fashion.
[0192] The projection device splits the radiation spectrum into
several partial light bundles R.sub.1G.sub.1B.sub.1,
R.sub.2G.sub.2B.sub.2, . . . R.sub.NG.sub.NB.sub.N. Each bundle is
modulated by different image modulators, which could be one or more
DMDs. The beams are then reunited by a beam integrator and
projected onto a diffuser.
[0193] The shutter may comprise a switchable aperture array,
wherein each aperture has an interference filters such that only
one light bundle will be transmitted. For example, stripes 1, N+1,
2N+1 etc would pass light bundle R.sub.1G.sub.1B.sub.1, stripes 2,
N+2, 2N+2 etc would pass light fiom bundle R.sub.2G.sub.2B.sub.2,
and stripes N, 2N, 3N etc would pass light from bundle
R.sub.NG.sub.NB.sub.N. Each light bundle and its corresponding set
of slits will form an independent system, each system superimposed
such that they are operated in the same way as a known scanning
slit display. Variations of this method may be used in other 3D
display system, including static parallax barrier systems.
[0194] Combining with Lens System
[0195] One way to increase bandwidth and brightness for the display
without higher frame rate is to combine the technology with similar
principles to those used in lenticular displays. An example of this
is illustrated in FIG. 6. This would also be an improvement over
current lenticular displays, where the main problem is getting a
wide field of view and many views without making pixel size or
pixel count too challenging.
[0196] What is required is a lens or holographic optical element
which is placed upon the shutter, just before, just after or both.
There will be one lens or optical equivalent for each slit. It acts
such that from any point of the display there will be a cone going
out to the lens and then the lens will form this into a parallel
beam of light the same width as the lens or slit. For horizontal
parallax only it should simply act as transparent in the vertical
direction. Viewers sufficiently far away from the display will see
pixels the width of the lens with a colour that is the combination
of light from a section of the display. This is mainly an advantage
compared to having no lens if the resolution of the display is
higher than the resolution of the shutter. Hence, it is mainly of
interest as a way of increasing bandwidth when it is not possible
to increase the frame rate further. Also, it would improve the
brightness compared to increasing the frame rate.
[0197] Combining with Directional Diffuser
[0198] One way to increase bandwidth and brightness for the display
without higher frame rate is to combine the technology with similar
principles to those used in holographic diffuser displays.
[0199] In effect the directional diffuser, which is sometimes
called an asymmetric diffuser, allows three separate images to be
superimposed on each other. However, from any one viewing angle or
vantage point one should only see a single image. To the left in
the FIG. 7 one can see the normal setup for the scanning slit. More
or less parallel light comes in and is scattered in all directions
by the diffuser. Hence the open slits must be sufficiently spaced
apart to avoid cross talk between the images displayed for the
respective slits. With the directional diffuser on the right open
slits can be put closer together. This is because the cross talk
from adjacent areas on the diffuser ends up coming from a different
projector. Consider region b. With a normal diffuser this would be
synchronized with slit 8 and having slit 5 and 11 open at the same
time would give cross talk for large viewing angles. However, if
for such viewing angles one would see information from a different
projector the cross talk could contain images rendered for slit 5
and 11. These images would not be seen through slit 5 since the
angle looking into the diffuser would be different. This setup
requires a diffuser that can give a controlled scattering angle and
that the light zones from different projectors can be accurately
aligned. One way to achieve this is using a holographic optical
element.
[0200] In some instances it is not desirable to have multiple
projectors. For example one might prefer to have a smaller form
factor by increasing the bandwidth of a single projector. Time
multiplexing will always have an upper limit beyond which other
methods must be used to increase bandwidth. For a single projector
the next step is then to increase resolution and use this to
increase the number of views.
[0201] One option would be to use a wedge with a slightly altered
geometry. Consider the wedge in FIG. 8. In this case a projector
with very high horizontal resolution could be used to create a wide
display.
[0202] Now, instead of a wide wedge one could allow the light for
high viewing angles to reflect back on the central strip as in FIG.
9. This would actually be the equivalent of having three lower
resolution projectors projecting from three different positions.
The actual projector projecting fiom straight on (section B above)
and two apparent projectors projecting from the sides (sections A
and C above). By combining this principle with a directional
diffuser one could increase bandwidth using a single high
resolution high frame rate projector. This would be similar to
having three projectors projecting the image areas A, B and C
respectively from the three different projector positions shown in
FIG. 9. The directional diffuser will ensure that from any one
viewing angle one would only see one of the images A, B or C. A
similar effect can be achieved by using mirrors in an optical
system not using a wedge. The light path would then be open from
the projector to the diffusive screen and mirrors on the side of
the light path would create the reflective sides.
[0203] Optimising Bandwidth Use Through Head Tracking
[0204] The 3D image quality can be improved by directing the same
bandwidth to a narrower field of view. One way to do this is to use
one or more head tracking devices. Such devices can locate where
one or more viewers are located in relation to the display. This
information can be used to produce viewing cones centred on the
position of each viewer. When the viewers move, the centres of the
viewing cones are moved too. The viewing cones can contain more
than two views of the scene and be wider than the distance between
the observer's eyes. This way the eye tracking system does not need
to be as accurate as for existing eye tracking displays.
[0205] Eye tracking can also be used to identify which part of a
scene the user is focusing on. Because the image quality of the
scene varies with distance to the central image plane it can in
some situations be desirable to shift the depth plane according to
where the user is focusing. Hence, the area in focus can be placed
as close to the central image plane as possible. The functionality
can be implemented in either hardware or software. One way to
implement this depth-of-field effect in software is to accumulate
multiple renders of a scene from slightly different perspectives,
ensuring that the camera frustums all intersect at the central
image plane.
[0206] Optimising Bandwidth Use Through Multiple Depths
[0207] The requirement for high bandwidth is generally more
important for scenes that are very deep, i.e. where there are
objects spread over a large depth. For example, this could be a
problem when there is a background far behind the main scene. One
way to improve backgrounds is to have more than one display that
shows an image. In FIG. 10, Diffuser 1 would show the main image
that is synchronised with the shutter. This will be transparent for
light coming from Diffuser 2, and diffusive for light coming from
the projector. One way to achieve that effect is to use a
holographic optical element. Diffuser 2 will show background
information, i.e. objects behind Diffuser 2 such as Object 2 below.
Diffuser 1 will show all other information. If one would like to
avoid objects to appear as semi-transparent one could synchronise
both the image on Diffuser 1 and Diffuser 2 to ensure that for any
one viewing angle only one of the Diffusers will show information.
One could also place a second shutter behind Diffuser 1. This would
have pixels that are transparent when a pixel on Diffuser 1 is
supposed to be transparent and black in all other instances. In
this case Diffuser 2 could be an image source which is constant for
all frames and only needs to be updated at the animation rate of
the overall scene.
[0208] Effective Use of Bandwidth
[0209] The principle of sharing time space between subframes can be
extended even further. The effective resolution of a scanning slit
display system decreases with a virtual point's distance from the
diffuser/display plane. One remedy is to make slits narrower by
introducing black stripes between slits. In FIG. 11 odd slits would
always be closed and the even slits would be scanned.
[0210] This could give an acceptable image as it is how static
parallax barriers work. It would be dimmer though and the black
stripes could be irritating. The gain would be the narrower slits,
which would decrease the size of the volume swept by a display
pixel shown in FIG. 13.
[0211] Instead of blocking out every odd slit, that slit could show
the average value of adjacent slits. For example, slit 9 would show
pixel values that are an average between the subframes for slit 8
and 10. That could be achieved by opening the shutter in slit 9
half way into the subframe for slit 8 and close it half way into
the subframe for slit 10. See timing diagram in FIG. 12.
[0212] Another way of explaining it is that two adjacent slits will
always have a period when they are open at the same time.
[0213] For points on the display plane, the pixel values will be
identical for slits 8 and 10, and as a result for slit 9, assuming
there are no lighting effects. Hence brightness has increased and
the stripe has been removed compared to the setup in FIG. 11. For
points out of the display plane the pixel values on the diffuser
will be different between subframes 8 and 10. If one considers the
volume swept by the same pixel on the diffuser for slit 8 and 10,
one will see that there is a large overlap of these with that of
the same pixel for slit 9. As a result, the pixel value for frame 9
would have been highly dependent on the pixel value for slits 8 and
10 even if one could show a unique frame for slit 9. It seems like
the number of views have doubled. The compromise is that
transitions between adjacent views will be limited. For example, it
will not be possible for a pixel on the display to go from full
black to full white in one view or slit increment. Instead one may
be restricted to go from full black to 50% grey. It should be noted
however that this limitation may not cause significant visual
degradation of the scene. In order to understand this, consider
FIG. 13. It shows a pixel on the diffuser and an open slit in the
shutter. The cone represents the area in which a virtual object
should influence the state of the pixel on the diffuser for an
observer moving freely in front of the diffuser.
[0214] FIG. 14 represents the cones for the same pixel for two
adjacent shutter slits. What becomes clear is that there is
considerable overlap between the two areas. For example in the
plane of the diffuser the overlap will be total. It should be noted
that there is also considerable overlap at other depths as well,
though the overlap is not total so the pixel will in many instances
have different values for different shutter slits. For example, the
virtual Object 1 should only influence the pixel value for the open
slit. Object 2 on the other hand should influence the pixel value
for both slits examined.
[0215] For objects in or near the diffuser plane there will not be
large transitions for pixels on the diffuser between adjacent
frames (corresponding to adjacent shutter slits). This makes sense
since an image in this plane will look the same in all directions.
It is not necessary to have a 3D display to show such an image (if
one ignores lighting effects). For objects further away from the
diffuser they may or may not lead to large transitions in pixel
values depending on where they are placed, i.e. Object 1 would lead
to a larger transition than Object 2 would, since Object 2 is
present in the cones for both slits. This is an inherent
restriction in the display system, which is a reflection of the
fact that with limited bandwidth it is not possible to make an
ideal representation of reality. It should be noted however that
for many rendering methods the fact that the two cones above have a
large overlap will mean that the transition in pixel values will be
restricted. The conclusion is that the transition in the value that
the same pixel takes for two adjacent subframes is restricted by
the inherent limitations in the system geometry.
[0216] In fact, one could allow subframes to overlap even more to
give higher brightness or more views. The result would be a higher
interdependence, and interdependence not only with the adjacent
slits, but also with slits further away.
[0217] The scheme can give a more accurate interpolation by
ensuring that the bit sequencing for the time multiplexed display
is such that all bits are proportionally represented in each time
window where two or more adjacent shutters are open simultaneously.
The example in FIG. 15 shows one such bit sequence for a 3 bit
frame. The LSB+1 and the MSB are split in two parts on either side
of the half way point in the frame. The LSB is not split, but is
placed in the centre of the subframe.
[0218] A further extension would involve a shutter with pixels or
other apertures rather than slits. In this case there could be
overlap in time both in the horizontal and vertical direction.
[0219] In some instances the system could also be improved by
analysing the similarity between subsequent subframes, either
locally on different parts of the display or the whole display. The
principle would be the same for both whole and partial subframes.
The time overlap could then be adapted to the difference between
subsequent frames. The order of the subframes could be changed such
that the sum of differences between frames is minimised or such
that the maximum difference is minimised or the average difference
is minimised or some other quantitative measure.
[0220] Varying Slit Width Within 3D Frame
[0221] As mentioned in the previous section, the effective
resolution of a scanning slit display system decreases with a
virtual point's distance from the diffuser/display plane, and one
way to reduce this effect is to reduce shutter slit width. However,
the requirement to have thin slits is typically more important for
virtual points far away from the diffuser plane than for those
close to it. At the same time it may be acceptable to have lower
image quality for virtual points far away from the display. To take
advantage of this fact one can construct a system that effectively
makes up two or more interlaced systems, each with a different slit
width.
[0222] Consider a basic setup for a simple system consisting of 6
unique frames and 6 slit groups described in FIG. 17. Each subframe
is shown within a frame cycle as shown in FIG. 19. The cycle is
repeated at a rate that is sufficiently fast for a viewer not to
perceive flicker. Because of this fact, the cycle can be changed
without causing flicker. For example, the first half of each
subframe could be placed at the start of the cycle and the second
half at the end of the cycle as illustrated in FIG. 20. This
requires that the shutter sequence changes to match the new partial
subframes. As an example, the first half of the subframe could
contain the MSB and the second half lower order bits.
[0223] If the shutter sequence is changed further it is possible to
have narrower slits for the first sets of subframes than for the
second sets of subframes. FIG. 21 shows such an example where the
cycle consists of a set of 9 subframes scanned with 9 slit groups,
and a second set of 3 subframes scanned with 3 slit groups. In this
example, the slit width for the second set of subframes is three
times wider than the slit width for the first set of subframes.
This way a system with narrow slits and better depth properties has
been superimposed with a system of wider slits. Compared to the
base example part of the system has narrower slits and part of the
system has wider slits. The light reaching the eye from any one
pixel will be the same. However, if one had reduced the slit width
for the entire system to fit in 9 unique frames, the light output
would need to be reduced to maintain the same cycle length.
[0224] The above is only an example. The system can be split into
any number of subframes and the duration of each subframe can be
different. The order of subframes within a cycle can also be
changed. The method can be applied even in a system without field
sequential colour.
[0225] It should also be noted that the slits do not need to be
physically wider or narrower. Instead the same effect can be
achieved by switching one, two or more groups of slits
simultaneously.
[0226] In some instances it is advantageous to have the first
subframes be multiple of the second set of subframes, such that the
information rendered for the first subframes can be used for the
second subframes. In the example above the multiple is 3 and as an
example partial frames 2, 5 and 7 from the first set of frames
could be used as the three subframes for the second set.
[0227] An extreme case of the method is to add a single subframe
within the cycle where the full shutter is transparent and a frame
or subframe displayed.
[0228] In some instances the method can be improved by only showing
data for parts of the virtual scene for a particular set of
subframes, and showing another part or the whole virtual scene for
another set of subframes.
[0229] The slit width can also be made to vary along the width of
the display. Depending on the scene one may wish to prioritise
different areas. For example, in scenes where the focus tends to be
on objects at the centre of the display the slits could be narrower
at the centre of the display than at the sides. The zone with
narrower slits could also be made to move dynamically. By using eye
tracking or another user device to change the zone, one can ensure
that slits are narrower in the part of the display where the user
is focusing.
[0230] Embodiments of the present invention have been described
with particular reference to the examples illustrated. However, it
will be appreciated that variations and modifications may be made
to the examples described without departing from the scope of the
present invention.
* * * * *