U.S. patent application number 14/486546 was filed with the patent office on 2015-01-01 for image processing device, method of controlling image processing device, and program for enabling computer to execute same method.
This patent application is currently assigned to Sony Corporation. The applicant listed for this patent is Sony Corporation. Invention is credited to Yasutaka HIRASAWA, Natsuki KANO.
Application Number | 20150003740 14/486546 |
Document ID | / |
Family ID | 47390766 |
Filed Date | 2015-01-01 |
United States Patent
Application |
20150003740 |
Kind Code |
A1 |
KANO; Natsuki ; et
al. |
January 1, 2015 |
IMAGE PROCESSING DEVICE, METHOD OF CONTROLLING IMAGE PROCESSING
DEVICE, AND PROGRAM FOR ENABLING COMPUTER TO EXECUTE SAME
METHOD
Abstract
There is provided an image processing device including a focus
area setting section, a depth acquisition section, and a blur
processing section. The focus area setting section sets each of a
plurality of areas as a focus area, and the each of the plurality
of areas is intended to be kept in focus in a captured image of a
subject. The depth acquisition section acquires a depth of the
subject in relation to each pixel in the image. The blur processing
section sets each of the pixels in the image as a target pixel and
performs a blur process on the target pixel in accordance with a
minimum value of absolute value of each depth difference between
depth corresponding to the target pixel and depth corresponding to
each of the focus area.
Inventors: |
KANO; Natsuki; (Tokyo,
JP) ; HIRASAWA; Yasutaka; (Tokyo, JP) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Sony Corporation |
Tokyo |
|
JP |
|
|
Assignee: |
Sony Corporation
Tokyo
JP
|
Family ID: |
47390766 |
Appl. No.: |
14/486546 |
Filed: |
September 15, 2014 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
13527905 |
Jun 20, 2012 |
8879847 |
|
|
14486546 |
|
|
|
|
Current U.S.
Class: |
382/195 ;
382/255 |
Current CPC
Class: |
G06T 2207/10028
20130101; G06T 2207/10004 20130101; G06T 5/002 20130101 |
Class at
Publication: |
382/195 ;
382/255 |
International
Class: |
G06T 5/00 20060101
G06T005/00 |
Foreign Application Data
Date |
Code |
Application Number |
Jun 28, 2011 |
JP |
2011-142807 |
Claims
1. An image processing device comprising: circuitry configured to
acquire a depth corresponding to each of a plurality of focus areas
in an image; and perform a blur process on a subject in the image
by changing a degree of a blur process on the subject in accordance
with the depth of a focus area located closest to subject or
located furthest from the subject.
2. The image processing device according to claim 1, wherein the
circuitry is configured to: extract a plurality of areas in the
image as candidates areas for the plurality of focus areas; and
select a plurality of candidate areas among the extracted candidate
areas and set each of the plurality of candidate areas as the
plurality of focus areas.
3. The image processing device according to claim 2, wherein the
circuitry is configured to recognize a captured area of a given
object in the image and extract the area as a candidate area.
4. The image processing device according to claim 2, wherein the
circuitry is configured to recognize a captured area of a face in
the image and extract the area as a candidate area.
5. The image processing device according to claim 2, wherein the
circuitry is configured to: decide a composition to define an
arrangement of the subject in the image; and extract a plurality of
areas as candidate areas based on the decided composition.
6. The image processing device according to claim 2, wherein the
circuitry is configured to: set a priority for the selection of the
candidate areas to the extracted candidate areas; and select the
candidate areas based on the set priority.
7. The image processing device according to claim 6, wherein the
circuitry is configured to: register a captured image of a given
subject; and set a priority to a candidate area based on a
similarity between the registered image and the candidate area.
8. The image processing device according to claim 6, wherein the
circuitry is configured to set the priority based on an area size
of the candidate area.
9. The image processing device according to claim 6, wherein the
circuitry is configured to set the priority based on the depth
corresponding to any one pixel in the candidate area.
10. The image processing device according to claim 9, wherein the
circuitry is configured to set the priority of a candidate area to
be higher as the depth corresponding to any one pixel in the
candidate area is closer to a maximum value or minimum value of the
obtained depth.
11. The image processing device according to claim 1, wherein the
image includes a criterion image and a reference image used in
acquiring the depths, and the circuitry is configured to detect a
distance between any one pixel in the criterion image and a pixel,
corresponding to the pixel in the criterion image, in the reference
image as a parallax; generate the depth in relation to the pixel in
the criterion image based on the detected parallax; and perform the
blur process on each of the pixels in the criterion image as the
target pixel.
12. The image processing device according to claim 1, wherein the
blur process is calculated by using the following expression:
.delta..sub.j=D.times.d.sub.j--min.times.R/d.sub.j where
.delta..sub.j is a degree of blur, D is an effective diameter of an
image capturing apparatus by which the image is captured, d.sub.j
is the depth corresponding to a target pixel, .DELTA.d.sub.j--min
is the minimum value, and R is an imaging magnification of the
image.
13. The image processing device according to claim 1, wherein the
circuitry is configured to set the focus areas in accordance with
an operation signal for specifying the focus areas.
14. The image processing device according to claim 1, wherein the
circuitry is configured to set the focus areas by limiting the
number of focus areas to be set to a value less than or equal to a
threshold.
15. A method of controlling an image processing device, comprising:
acquiring, by circuitry, a depth corresponding to each of a
plurality of focus areas in an image; and performing, by the
circuitry, a blur process on a subject in the image by changing a
degree of a blur process on the subject in accordance with the
depth of a focus area located closest to subject or located
furthest from the subject.
16. A non-transitory computer-readable medium including a program
for enabling a computer to: acquire a depth corresponding to each
of a plurality of focus areas in an image; and perform a blur
process on a subject in the image by changing a degree of a blur
process on the subject in accordance with the depth of a focus area
located closest to subject or located furthest from the subject.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application is a continuation of and is based upon and
claims the benefit of priority under 35 U.S.C. .sctn.120 for U.S.
Ser. No. 13/527,905, filed Jun. 20, 2012, and claims the benefit of
priority under 35 U.S.C. .sctn.119 from Japanese Patent Application
No. 2011-142807, filed Jun. 28, 2011. The entire contents of which
are incorporated herein by reference.
BACKGROUND
[0002] The present disclosure relates to an image processing
device, a method of controlling the image processing device, and a
program for enabling a computer to execute the method. More
particularly, the present disclosure relates to an image processing
device capable of performing a blur process, a method of
controlling the image processing device, and a program for enabling
a computer to execute the method.
[0003] In recent years, an image capturing apparatus which can
measure a depth of a target to be captured has been commonly used.
An example of a method for measuring the depth includes a depth
from defocus (DFD) method, a stereo matching method, and so on. In
the DFD method, the degree of blur for a subject in an image is
analyzed, and the depth from image capturing apparatus to the
subject is measured based on a relationship between the degree of
blur and the depth. In the stereo matching method, a parallax is
detected from a criterion image and a reference image, and the
depth from the image capturing apparatus to the subject is measured
based on a relationship between the parallax and the depth.
[0004] There has been proposed an image capturing apparatus which
measures the depth from the image capturing apparatus to a subject
using the above-mentioned methods and performs a blur process on a
background of a main subject in accordance with the depth in a
state that the main subject is kept in focus (see, for example,
Japanese Patent Application Laid-open Publication No. 2003-37767).
An image with enhanced perspective is generated by performing the
blur process in accordance with the depth. This blur process is
often used for clearly showing a person against a background in
portrait photographs. In the image capturing apparatus equipped
with image capturing elements having small light-receiving area, an
image with a relatively less blurring is captured because of the
characteristics of the image capturing elements, thus this blur
process is commonly used to enhance the perspective of an
image.
SUMMARY
[0005] However, in the above-described related art, when two or
more areas are set to be kept in focus during the blur process (the
areas are hereinafter referred to as "focus area"), it may be
incapable of representing a natural perspective. As an example, it
is assumed that two areas of close and far subjects are set as a
focus area and the background is blurred. When the background is
blurred in accordance with a distance from the close subject in the
depth direction, a boundary between the un-blurred far subject and
the blurred background becomes unnatural. On the other hand, when
the background is blurred in accordance with a distance from the
far subject, a boundary between the un-blurred close subject and
the blurred background becomes unnatural. As a result, when setting
a plurality of focus areas, there has been a problem that it may be
not possible to represent a natural perspective.
[0006] The present disclosure has been made in view of the above
situation. It is desirable to provide a technique for generating an
image capable of representing a more natural perspective even when
setting a plurality of focus areas.
[0007] According to an embodiment of the present disclosure, there
is provided an image processing device, a method of controlling the
image processing device, and a program for enabling a computer to
execute the method. The image processing device includes a focus
area setting section for setting each of a plurality of areas that
are to be kept in focus in a captured image of a subject as a focus
area, a depth acquisition section for acquiring a depth of the
subject in relation to each pixel in the image, and a blur
processing section for setting each pixel in the image as a target
pixel and for performing a blur process on the target pixel in
accordance with a minimum value of each depth difference absolute
value between depth corresponding to the target pixel and depth
corresponding to each of the focus area. According to this, it is
possible to perform the blur process on the target image in
accordance with a minimum value of each depth difference absolute
value between depth corresponding to the target pixel and depth
corresponding to each of a plurality of focus areas.
[0008] In the embodiment of the present disclosure, the focus area
setting section may include a candidate area extraction part for
extracting each of a plurality of areas to be a candidate for the
focus area in the image as a candidate area, and a candidate area
selection part for selecting a plurality of candidate areas among
the extracted candidate areas and for setting each of the plurality
of candidate areas as the focus area. Thus, each of the plurality
of candidate areas which are selected among the extracted candidate
areas can be set as the focus area.
[0009] In the embodiment of the present disclosure, the focus area
setting section may recognize a captured area of a given object in
the image and extract the area as the candidate area. Therefore,
the captured area of the object can be extracted as the candidate
area.
[0010] In the embodiment of the present disclosure, the candidate
area extraction part may recognize a captured area of a face in the
image and extract the area as the candidate area. As a result, the
captured area of the face can be extracted as the candidate
area.
[0011] In the embodiment of the present disclosure, the candidate
area extraction part may include a composition decision portion for
deciding a composition to define an arrangement of the subject in
the image, and a candidate area extraction processing portion for
extracting each of a plurality of areas as the candidate area based
on the decided composition. Consequently, the candidate area can be
extracted based on the decided composition.
[0012] In the embodiment of the present disclosure, the focus area
setting section may further include a priority setting part for
setting a priority for the selection of the candidate area to the
extracted candidate area, and the candidate area selection part may
select the candidate area based on the set priority. So, the
candidate area can be selected based on the set priority.
[0013] In the embodiment of the present disclosure, the priority
setting part may register a captured image of a given subject and
set the priority to the candidate area based on a similarity
between the registered image and the candidate area. Hence, the
priority can be set based on the similarity between the registered
image and the candidate area.
[0014] In the embodiment of the present disclosure, the priority
setting part may set the priority based on an area size of the
candidate area. According to this, the priority can be set based on
the area size of the candidate area.
[0015] In the embodiment of the present disclosure, the priority
setting part may set the priority based on the depth corresponding
to any one pixel in the candidate area. Thus, the priority can be
set based on the depth corresponding to any one pixel in the
candidate area.
[0016] In the embodiment of the present disclosure, the priority
setting part may set the priority of the candidate area to be
higher as the depth corresponding to any one pixel in the candidate
area is closer to a maximum or minimum value of the obtained depth.
Therefore, the priority of the candidate area can be set to be
higher as the depth of the candidate area is closer to a maximum or
minimum value of the obtained depth.
[0017] In the embodiment of the present disclosure, the image may
include criterion and reference images for using in the acquisition
of the depth. The depth acquisition section may include a parallax
detection part for detecting a distance between any one pixel in
the criterion image and a pixel in the reference image
corresponding to the pixel in the criterion image and for setting
the distance as a parallax, and a depth generation part for
generating the depth in relation to the pixel in the criterion
image based on the detected parallax. The blur processing section
may perform the blur process on each of the pixels in the criterion
image as the target pixel. As a result, the depth can be generated
based on the detected depth.
[0018] In the embodiment of the present disclosure, the blur
processing section may perform the blur process on the target
pixel, and the blur process may be calculated using the following
expression:
.delta..sub.j=D.times..DELTA.d.sub.j--min.times.R/d.sub.j
where .delta..sub.j is a degree of blur, D is an effective diameter
of an image capturing apparatus by which the image is captured,
d.sub.j is the depth corresponding to the target pixel,
.DELTA.d.sub.j--min is the minimum value, and R is an imaging
magnification of the image. According to this, the degree of blur
can be calculated from the above-mentioned expression.
[0019] In the embodiment of the present disclosure, the focus area
setting section may set the focus area in accordance with an
operation signal for specifying the focus area. Thus, the focus
area can be set in accordance with the operation signal.
[0020] In the embodiment of the present disclosure, the focus area
setting section may set the focus area by limiting the number of
focus areas to be set to a value less than or equal to a threshold.
Therefore, the number of focus areas to be set can be limited to a
value less than or equal to a threshold.
[0021] According to the embodiments of the present disclosure
described above, it is possible to generate an image capable of
representing a more natural perspective when setting a plurality of
focus areas.
BRIEF DESCRIPTION OF THE DRAWINGS
[0022] FIG. 1 is a block diagram illustrating a configuration
example of an image capturing apparatus according to a first
embodiment;
[0023] FIG. 2 is a diagram illustrating examples of image data and
depth information according to the first embodiment;
[0024] FIG. 3 is a diagram explaining a method for setting focus
area according to the first embodiment;
[0025] FIG. 4 is a diagram illustrating examples of depth
information, a measured depth, and a depth difference absolute
value according to the first embodiment;
[0026] FIG. 5 is a block diagram illustrating a configuration
example of an image processing device according to the first
embodiment;
[0027] FIG. 6 is a diagram illustrating an example of a depth
difference absolute value according to the first embodiment;
[0028] FIG. 7 is a diagram illustrating an example of a filter
order according to the first embodiment;
[0029] FIG. 8 is a diagram illustrating an example of an image data
after blurring according to the first embodiment;
[0030] FIG. 9 is a flowchart illustrating an operation example of
an image processing device according to the first embodiment;
[0031] FIG. 10 is a flowchart illustrating a setting process of
focus area according to the first embodiment;
[0032] FIG. 11 is a flowchart illustrating an example of a blur
process according to the first embodiment;
[0033] FIG. 12 is a block diagram illustrating a configuration
example of an image capturing apparatus according to a second
embodiment;
[0034] FIG. 13 is a block diagram illustrating a configuration
example of an image processing device according to the second
embodiment;
[0035] FIG. 14 is a block diagram illustrating a configuration
example of a depth information acquisition section according to the
second embodiment;
[0036] FIG. 15 is a diagram illustration a relationship between a
parallax and a depth according to the second embodiment;
[0037] FIG. 16 is a flowchart illustrating an operation example of
the image processing device according to the second embodiment;
[0038] FIG. 17 is a block diagram illustrating a configuration
example of a focus area setting section according to a third
embodiment;
[0039] FIG. 18 is a flowchart illustrating an example of a focus
area setting process according to the third embodiment;
[0040] FIG. 19 is a block diagram illustrating a configuration
example of a focus area setting section according to a fourth
embodiment;
[0041] FIG. 20 is a diagram illustrating examples of image data
before and after blurring according to the fourth embodiment;
[0042] FIG. 21 is a flowchart illustrating an example of a focus
area setting process according to the fourth embodiment;
[0043] FIG. 22 is a block diagram illustrating a configuration
example of a focus area setting section according to a fifth
embodiment;
[0044] FIG. 23 illustrates examples of image data before and after
blurring when the rule of thirds composition is set according to
the fifth embodiment;
[0045] FIG. 24 illustrates examples of focus area information and
image data before and after blurring when pattern composition is
set according to the fifth embodiment; and
[0046] FIG. 25 is a flowchart illustrating an example of a focus
area setting process according to the fifth embodiment.
DETAILED DESCRIPTION OF THE EMBODIMENT(S)
[0047] Hereinafter, preferred embodiments of the present disclosure
will be described in detail with reference to the appended
drawings. Note that, in this specification and the appended
drawings, structural elements that have substantially the same
function and structure are denoted with the same reference
numerals, and repeated explanation of these structural elements is
omitted.
[0048] An embodiment for implementing the present disclosure
(hereinafter simply referred to as "embodiment") will be described
below. In addition, the description will be made in the following
order. [0049] 1. First Embodiment (image process: an example of
setting a plurality of focus areas) [0050] 2. Second Embodiment
(image process: an example of using the stereo matching) [0051] 3.
Third Embodiment (image process: an example of performing an object
recognition) [0052] 4. Fourth Embodiment (image process: an example
of performing a facial recognition) [0053] 5. Fifth Embodiment
(image process: an example of selecting a focus area based on the
composition)
1. First Embodiment
Configuration Example of Image Capturing Apparatus
[0054] FIG. 1 is a block diagram illustrating a configuration
example of an image capturing apparatus 100 according to a first
embodiment of the present disclosure. The image capturing apparatus
100 captures a subject and performs a blur process on an image of
the captured subject. The image capturing apparatus 100 includes an
image capturing unit 150, an image data storage unit 200, a depth
measurement unit 250, an operation unit 300, an image processing
device 400, and a display unit 500.
[0055] The image capturing unit 150 captures a subject and
generates image data. The image capturing unit 150 provides the
image data storage unit 200 with the generated image data through a
signal line 155.
[0056] The image data storage unit 200 stores the image data. More
specifically, the image data storage unit 200 stores the image data
which is generated by the image capturing unit 150 and the image
processing device 400.
[0057] The depth measurement unit 250 measures the depth of a
subject (a distance from the image capturing apparatus to the
subject). For example, the depth measurement unit 250 analyzes a
degree of blur of the subject in an image and measures the depth
using the DFD method in which the depth of the subject is obtained,
on the basis of a relationship between the degree of blur and the
depth of the subject. The depth measurement unit 250 measures the
depth in relation to each pixel in the image and generates depth
information indicating the depth. The depth measurement unit 250
provides the image processing device 400 with the generated depth
information through a signal line 251. The depth measurement unit
250 also can measure a depth using any measurement method other
than the DFD method. For example, the depth measurement unit 250
can use a laser measurement method which irradiates a laser light
onto a subject and measures a depth based on a delay time of the
reflected light with respect to the irradiation time.
[0058] The operation unit 300 detects an operation signal for
operating the image capturing apparatus 100. The operation signal
includes an operation signal for setting a plurality of focus areas
in an image. The operation unit 300 provides the image processing
device 400 with the detected operation signal through a signal line
301.
[0059] The image processing device 400 generates the image data
with enhanced perspective by performing a blur process on the image
data. More specifically, the image processing device 400 sets a
plurality of focus areas in an image in accordance with the
operation signal. The image processing device 400 also acquires
depth information generated by the depth measurement unit 250. In
the image processing device 400, each pixel in an image is set as a
target pixel P.sub.j, and each depth difference absolute value
between depth of the target pixel and depth of each of the
plurality of focus areas is calculated as a depth difference
absolute value .DELTA.d.sub.j--fm for each target pixel. Here, j
denotes a sequential number used for identifying each of the pixels
in the image and is an integer from 0 to N-1. N is a value
indicating the number of total pixels in the image and thus becomes
a value obtained by multiplying the number of pixels in the
vertical direction to the number of pixels in the horizontal
direction. m is a sequential number used for identifying each of
the focus areas and is an integer from 0 to M-1. M is the number of
the set focus areas and is an integer greater than or equal to 2.
Therefore, when M focus areas are set, M depth difference absolute
values are calculated for each pixel in the image.
[0060] After M depth difference absolute values are calculated for
each target pixel, the image processing device 400 selects a depth
difference absolute value as a depth difference minimum value
.DELTA.d.sub.j--min for each target pixel. Thus, if the total
number of pixels in the image is N, the total N depth difference
minimum value are selected.
[0061] After the depth difference minimum value .DELTA.d.sub.j--min
is selected, the image processing device 400 calculates the degree
of blur in accordance with the depth difference minimum value
.DELTA.d.sub.j--min for each target pixel.
[0062] The image processing device 400 performs the blur process on
the image data in accordance with the calculated filter order. The
image processing device 400 provides the image data storage unit
200 with the image data on which the blur process is performed,
through a signal line 405.
[0063] The display unit 500 reads out image data from the image
data storage unit 200 and displays the image data.
[0064] FIG. 2 is a diagram illustrating examples of image data and
depth information according to the first embodiment. FIG. 2(a)
illustrates an example of image data 600 which is inputted to the
image processing device 400. In the image data 600, a subject 601
containing a car and a driver, subjects of trees 602 and 603, and a
subject of house 604 are captured respectively.
[0065] FIG. 2(b) is a diagram illustrating an example of depth
information 610 obtained in synchronization with the image data
600. The depth information 610 is data which indicates the depth of
a subject corresponding to each pixel, and the depth is represented
with a pixel value. For example, in the depth information 610, a
luminance value is set by using a pixel value indicating the depth,
and the luminance value is set to be lower as the depth is
increased. The depth information 610 includes depth information of
each of subjects 611, 612, 613 and 614 which respectively
correspond to the subjects 601, 602, 603 and 604 in the image data
600. It is assumed that the subject 611 has the minimum depth among
the subjects 611 to 614 and their respective backgrounds.
Subsequently, it is assumed that depths of the remaining subjects
and their respective backgrounds are increased in the order of the
subject 612, the subject 613, the subject 614 and the backgrounds.
In this case, in the depth information 610, the luminance value of
the foremost subject 611 is set to be the lowest. The luminance
values of the subjects and their respective backgrounds are set to
be increased in the order of the subject 612, the subject 613, the
subject 614 and the backgrounds in accordance with their respective
depths. Alternatively, in the depth information 610, the luminance
value may be set to be higher as the depth is decreased.
[0066] FIG. 3 is a diagram explaining a method for setting a focus
area according to the first embodiment. As an example, the image
processing apparatus 100 displays the image data 600 on a touch
panel and receives user operations. When any one pixel of the image
data 600 is designated by touching the pixel with user's finger,
the image processing device 400 of the image capturing apparatus
100 detects a subject area containing the pixel. For example, the
image processing device 400 detects the subject area containing the
designated pixel by analyzing the distribution of luminance
information in the image. The image processing device 400 may be
configured to detect the subject area by analyzing the distribution
of color information in addition to the distribution of the
luminance information. The image processing device 400 sets the
detected subject area as a focus area. The image capturing
apparatus 100 may be configured to set the focus area by entering
the coordinates of the pixel using numeric keys by a user instead
of using the touch panel.
[0067] As an example, it is assumed that a pixel 605 in the subject
601 and a pixel 606 in the subject 604 are designated. In this
case, the image processing device 400 detects the subjects 601 and
604 containing the designated pixels and sets each of the subjects
as a focus area.
[0068] FIG. 4 is a diagram illustrating examples of depth
information, a measured depth, and depth difference absolute value
according to the first embodiment. When the focus area is set, the
image processing device 400 designates each pixel in the image as a
target pixel and calculates a depth difference absolute value by
using a depth of the target pixel and each depth of the plurality
of focus areas. For example, a depth difference absolute value of a
target pixel P.sub.j in the subject 612 may be calculated. In this
case, when the subjects 601 and 604 are each set as the focus area,
the image processing device 400, as shown in FIG. 4(a), acquires a
depth d.sub.j of the target pixel P.sub.j and the depths d_f0 and
d_f1 which each correspond to any one pixel in the focus area. The
image processing device 400 calculates the depth difference
absolute values .DELTA.d.sub.j--f0 and .DELTA.d.sub.j--f1 from the
following Expressions (1) and (2):
.DELTA.d.sub.j--f0=|d.sub.j-d.sub.--f0| (1)
.DELTA.d.sub.j--f1=|d.sub.j-d.sub.--f1| (2)
[0069] The image processing device 400 calculates the degree of
blur .delta..sub.j for the target pixel P.sub.j from the depth
difference absolute values obtained by Expressions (1) and (2).
[0070] FIG. 4(b) is a diagram illustrating an example of positional
relationships in the depth direction of the subject in the image
shown FIG. 4(a). More specifically, in a landscape shown in FIG.
4(a), in a case where x-axis is parallel to the horizontal
direction, y axis is parallel to the vertical direction, and z-axis
is parallel to the depth direction, FIG. 4(b) is the landscape seen
from the direction of the x-axis. As shown in FIG. 4(b), when the
focus area f0 is set to the subject 611, the distance on the z-axis
from the image capturing apparatus 100 to the subject 611 becomes
the depth d_f0. When the focus area f1 is set to the subject 614,
the distance on the z-axis from the image capturing apparatus 100
to the subject 614 becomes the depth d_f1. When a pixel on the
subject 612 is set to the target pixel P.sub.j, the distance from
the image capturing apparatus 100 to the subject 612 becomes the
depth d.sub.j. In this case, the depth difference absolute value
.DELTA.d.sub.j--f0 is a distance from a side on the subject 611
facing the image capturing apparatus 100 to a side on the subject
612 facing the image capturing apparatus 100. The depth difference
absolute value .DELTA.d.sub.j--f1 is a distance from a side on the
subject 612 facing the image capturing apparatus 100 to a side on
the subject 614 facing the image capturing apparatus 100.
[0071] When each side of the focus areas on the subjects 611 and
614 facing the image capturing apparatus 100 are not flat, it will
actually cause some deviation in the depth of each pixel in the
focus area. However, the degree of deviation is usually negligibly
small, considering the measurement unit of the depth. For this
reason, the image processing device 400 can acquire the depth
corresponding to any pixel in the focus area (e.g., designated
pixels 605 and 606) as the depth corresponding to the focus
area.
[0072] Configuration Example of Image Processing Device
[0073] FIG. 5 is a block diagram illustrating a configuration
example of the image processing device 400 according to the first
embodiment. The image processing device 400 includes a focus area
setting section 410, a depth information acquisition section 480,
and a blur processing section 490.
[0074] The focus area setting section 410 is configured to set a
plurality of focus areas f0 to f(M-1) in accordance with operation
signals. The focus area setting section 410 provides the blur
processing section 490 with focus area information indicating the
set focus area through a signal line 415. The focus area
information may include the coordinates of pixel group forming the
contour of each focus area. The focus area information may include
the coordinates of all pixels in the focus area. The focus area
information may be image data in which each pixel value in the
focus area is set to one (e.g., "1") of binary values and each
pixel value in the region other than the focus area is set to the
other (e.g., "0") of the binary values. Identification information
for identifying each of the set focus areas is added to this image
data. The identification information may include the coordinates of
pixel in the image designated by a user and the sequential number
(0 to M-1) corresponding to the coordinates.
[0075] The focus area setting section 410 detects the areas of the
subject containing the designated pixel, and moreover the focus
area setting section 410 may detect the subject by using other
methods besides the above-mentioned method. For example, the focus
area setting section 410 may detect a group of pixels having the
depth in which the difference between the depth of the pixel in the
group and the depth corresponding to the designated pixel is less
than or equal to the predetermined value, with reference to the
depth information. And then, in the focus area setting section 410,
the detected group of pixels can be regarded as the subject area.
The focus area setting section 410 also can detect the subject by
performing edge detection and by extracting the contour of the
subject containing the designated pixel.
[0076] The focus area setting section 410 sets the detected subject
area as the focus area, and moreover the focus area setting section
410 may set an area in a figure (e.g., a circle) which has a given
size and contains the designated pixel, as the focus area.
[0077] In the focus area setting section 410, the number of the
focus areas may be limited to a value less than or equal to a
threshold T (T is an integer greater than or equal to 2). In this
case, when the number of the focus areas is equal to the threshold
T, the focus area setting section 410 is configured not to accept
operations for designating the subsequent focus areas.
[0078] The depth information acquisition section 480 acquires the
depth information generated by the depth measurement unit 250. The
depth information acquisition section 480 provides the blur
processing section 490 with the acquired depth information through
a signal line 485. The depth information acquisition section 480 is
an example of the depth acquisition section recited in claims.
[0079] The blur processing section 490 performs the blur process on
the image data. The blur processing section 490 includes a depth
difference absolute value calculation part 491, a depth difference
minimum value selection part 492, a filter order calculation part
493, and a blur filter 494.
[0080] The depth difference absolute value calculation part 491
calculates the depth difference absolute values .DELTA.d.sub.j--f0
to .DELTA.d.sub.j--f(M-1) for each target pixel P.sub.j (j ranges
from 0 to N-1) in the image based on focus area information and
depth information. The depth difference absolute value calculation
part 491 provides the depth difference minimum value selection part
492 with the calculated depth difference absolute value.
[0081] The depth difference minimum value selection part 492
selects the depth difference minimum value .DELTA.d.sub.j--min for
each target pixel in the image. The depth difference minimum value
selection part 492 provides the filter order calculation part 493
with the selected depth difference minimum value.
[0082] The filter order calculation part 493 calculates the filter
order .delta..sub.j for each target pixel P.sub.j in the image. The
filter order calculation part 493 may calculate the degree of blur
from the following Expression (3). However, in the following
Expression (3), ".delta..sub.j" denotes a degree of blur on the
target pixel P.sub.j, for example, a filter order of the blur
filter. An amount of blurring is increased or decreased in
accordance with the set filter order. ".DELTA.d.sub.j--min" denotes
a depth difference minimum value of the target pixel P.sub.j, and
the unit thereof is, for example, [m]. "A" denotes a coefficient
for converting a value of .DELTA.d.sub.j--min into
.delta..sub.j.
.DELTA.d.sub.j=Ad.sub.j--min (3)
[0083] The filter order calculation part 493 can calculate the
degree of blur .delta..sub.j using the following Expression (4),
instead of Expression (3). However, in the Expression (4), "D"
denotes an effective aperture of lens in the image capturing
apparatus 100, and the unit thereof is, for example, [mm].
"d.sub.1" denotes depth of the target pixel P.sub.j. "R" denotes an
imaging magnification of the image data. "C" denotes coefficients
for converting the amount of blurring obtained from D,
.DELTA.d.sub.j--min, d.sub.j, and R into the filter order
.delta..sub.j. The image processing device 400 calculates the
filter order for all pixels in the image using Expression (3) or
the following expression. When the following expression is used,
the amount of blurring to be adapted to characteristics of actual
optical systems is calculated. The filter order calculation part
493 provides the blur filter 494 with the calculated filter order
.delta..sub.j.
.delta. j = C D .DELTA. d j _min d j R ( 4 ) ##EQU00001##
[0084] The blur filter 494 performs the blur process on the image
data in accordance with the filter order .delta..sub.j. As an
example, the blur filter is a Gaussian filter which is a type of a
smoothing filter, and calculates the pixel value I'(x,y) of the
pixel after blurring (x,y) from the following expression. However,
in the following Expression (5), "I(x+k,y+l)" denotes the pixel
value of the pixel before blurring (x+k,y+l). "r" denotes a radius
of the Gaussian filter and is an integer greater than or equal to
0. "w(k,l)" denotes a weighting coefficient to be multiplied by the
pixel value I(x+k,y+1). In the following Expression (6), "a"
denotes the standard deviation, and a given real number is assigned
as the standard deviation. Considering the following Expressions
(5) and (6), the weighting coefficient is set to be higher as it is
closer to the center of the Gaussian filter and it is set to be
lower as it is closer to the peripheral of the Gaussian filter. In
the following Expression (7), "round( )" denotes a function which
performs a given rounding process for the numerical value in the
parentheses and returns an integer value greater than 0. For
example, the calculation which is rounded to the decimal point is
executed as the rounding process.
I ' ( x , y ) = k = - r r l = r r w ( k , l ) I ( x + k , y + 1 ) (
5 ) w ( x , y ) = 1 2 .pi..sigma. 2 exp { - ( x 2 + y 2 ) 2 .sigma.
2 } ( 6 ) r = round ( .delta. j / 2 ) ( 7 ) ##EQU00002##
[0085] The blur processing section 490 calculates the filter order
of the Gaussian filter as the degree of blur, and moreover the blur
processing section 490 may calculate parameters other than the
filter order as the degree of blur. For example, the blur
processing section 490 can calculate the standard deviation a as
the degree of blur.
[0086] The blur processing section 490 performs the blur process
using the Gaussian filter, and moreover the blur processing section
490 may use other blur filters such as a moving average filter.
[0087] Image processing device 400 is provided within the image
capturing apparatus 100. Alternatively, the image processing device
400 may be provided outside of the image capturing apparatus 100,
or the image processing device 400 and the image capturing
apparatus 100 may be separate components.
[0088] FIG. 6 is a diagram illustrating an example of a depth
difference absolute value according to the first embodiment. It may
be considered that the plurality of focus areas including focus
areas f0, f1 and f2 are set. It may be assumed that the respective
depths corresponding to the focus areas f0, f1 and f2 are 10 m, 30
m and 20 m. If the depth corresponding to pixel P.sub.0(0,0) is 12
m, then the depth difference absolute value representing the depth
difference between the depth of the pixel P.sub.0 and the depth of
the focus area f0 is 2 m. Also, the depth difference absolute
values between the depth of the pixel P.sub.0 and the depth of the
focus area f1 and between the depth of the pixel P.sub.0 and the
depth of the focus area f2 are 18 m and 8 m, respectively. The
lowest value of these depth difference absolute values is 2 m, and
thus 2 m is selected as a depth difference minimum value in the
pixel P.sub.0.
[0089] In a similar manner, if the depth corresponding to the pixel
P.sub.100(100,0) is 16 m, then 4 m is selected as the depth
difference minimum value in the pixel. If the depth corresponding
to the pixel P.sub.200(200,0) is 36 m, then 6 m is selected as the
depth difference minimum value in the pixel. If the depth
corresponding to the pixel P.sub.N-1(W-1,H-1) is 10.2 m, then 0.2 m
is selected as the depth difference minimum value in the pixel. W
denotes the total number of pixels in the horizontal direction in
the image, and is an integer greater than or equal to 1. H denotes
the total number of pixels in the vertical direction in the image,
and is an integer greater than or equal to 1. Therefore, the depth
difference absolute value is calculated for each of the set focus
areas of pixels in the image. The smallest value of the depth
difference absolute values is selected as the depth difference
minimum value for each pixel.
[0090] FIG. 7 is diagram illustrating an example of a filter order
according to the first embodiment. In the above-mentioned
Expression (3), it is assumed that "2" is assigned to the
coefficient A. In this case, if the depth difference minimum value
of the pixel P.sub.0 is 2 m, then "4" is calculated as the filter
order by the Expression (3). In a similar manner, if the depth
difference minimum values of the pixel P.sub.100 and P.sub.200 are
4 m and 6 m respectively, then "8" and "12" are each calculated as
the filter order by the Expression (3). If the depth difference
minimum value of the pixel P.sub.N-1 is 0.2 m, then "0.4" is
calculated as the filter order by the Expression (3).
[0091] FIG. 8 is a diagram illustrating an example of image data
before blurring 620 according to the first embodiment. The subjects
621 to 624 of the image data 620 correspond to the subjects 601 to
604 of the image data 600, respectively. Each of the subjects 621
to 624 is assumed to be set as a focus area. It is assumed that the
subject 621 has a minimum depth among the subjects 621 to 624.
Also, it is assumed that the depths of other subjects 622 to 624
are increased in the order of the subject 622, the subject 623, and
the subject 624. Also, it is assumed that the subjects 622 and 623
are located closer to the subject 621 than the subject 624, and the
subject 622 is located closer to the subject 621 than the subject
623.
[0092] In this case, the degree of blur is increased from a close
focus area (e.g., subject 621) to a far focus area in the depth
direction. The degree of blur becomes the largest at midpoint in
the depth direction of the close focus area (subject 621) and the
far focus area (subject 624). The degree of blur becomes decreased
as it is closer to the far focus area (subject 624) beyond the
midpoint depth. For example, since the subject 622 is located
closer to the subject 621 than that of the subject 623 in the depth
direction and the subject 622 is set as the focus area, the degree
of blur on the subject 622 becomes smaller than the subject 623.
Thus, even when a plurality of focus areas are set, the degree of
blur is changed in accordance with the depth from the focus area
located at the foremost point in the depth direction, thereby
obtaining a more natural perspective.
[0093] Operation Example of Image Processing Device
[0094] The operation example of the image processing device 400
according to the first embodiment will be described with reference
to FIG. 9 to FIG. 11. FIG. 9 is a flowchart illustrating an
operation example of the image processing device 400 according to
the first embodiment. This operation flow starts from entering the
captured image data into the image processing device 400. The image
processing device 400 performs a focus area setting process for
setting a plurality of focus areas (step S910). The image
processing device 400 acquires depth information from the depth
measurement unit 250 (step S950). The image processing device 400
performs a blur process for blurring the image data (step S960).
After performing step S960, the image processing device 400
terminates the operation for generating the image data after
blurring.
[0095] FIG. 10 is a flowchart illustrating an example of a focus
area setting process according to the first embodiment. The focus
area setting section 410 initializes variable i to zero (0) (step
S911) and receives an operation for setting a focus area (step
S912). The focus area setting section 410 determines whether or not
any one pixel is selected as a pixel in the focus area (step
S913).
[0096] When the focus area setting section 410 determines that any
one pixel is selected as a pixel in the focus area (step S913:
Yes), the focus area setting section 410 temporarily stores the
area containing the selected pixel as a candidate of focus area fi
(step S914). The focus area setting section 410 increments the
variable i by one (1) (step S915).
[0097] When determined that the pixel is not selected as a pixel in
the focus area (step S913: No) or after performing step S915, the
focus area setting section 410 determines whether or not the
selection of the pixel is confirmed (step S916). When determined
that the selection of the pixel is not confirmed (step S916: No),
the focus area setting section 410 returns the process to step
S913. When determined that the selection of the pixel is confirmed
(step S916: Yes), the focus area setting section 410 sets each
candidate as the focus area (step S917). After performing step
S917, the focus area setting section 410 terminates the focus area
setting process.
[0098] FIG. 11 is a flowchart illustrating an example of the blur
process according to the first embodiment. The blur processing
section 490 of the image processing device 400 acquires depths d_f0
to d_f(M-1) of all focus areas (step S961). The blur processing
section 490 initializes variables m and n to zero (0), and starts a
loop process. In this loop process, the blur processing section 490
performs steps S963 to S968 which will be described later, and
increments the variables in and n by 1, respectively. When m>W-1
and n>H-1 after incrementing the respective variables, the loop
process is terminated. And then, the process flow is returned to
step S963, and the loop process continues (step S962).
[0099] The blur processing section 490 sets pixel (m,n) to target
pixel P.sub.j (step S963). The blur processing section 490 acquires
depth d.sub.j of the target pixel P.sub.j (step S964). The blur
processing section 490 also calculates depth difference absolute
values .DELTA.d.sub.j--f0 to .DELTA.d.sub.j--f(M-1) from the depth
d.sub.j of the target pixel P.sub.j and the depths d_f0 to d_f(M-1)
of the focus area (step S965). The blur processing section 490
selects the minimum value of the calculated depth difference
absolute values .DELTA.d.sub.j--f0 to .DELTA.d.sub.j--f(M-1) as a
depth difference minimum value .DELTA.d.sub.j--min (step S966). The
blur processing section 490 calculates filter order .delta..sub.j
of the target pixel P.sub.j from the selected depth difference
minimum value .DELTA.d.sub.j--min using the above-mentioned
Expression (3) (step S967). The blur processing section 490
calculates the pixel value of the target pixel after blurring
P.sub.j using the above-mentioned Expressions (5) and (6) in
accordance with the calculated filter order (step S968).
[0100] After the loop process is completed (step S962), the blur
processing section 490 terminates the blur process.
[0101] As such, according to the first embodiment, the focus area
setting section 410 sets a plurality of focus areas. The depth
information acquisition section 480 acquires depths in relation to
pixels. The blur processing section 490 calculates each of the
depth difference absolute values between depth of pixels and depth
corresponding to each of the focus areas as a depth difference
absolute value for each pixel in the image. The blur processing
section 490 selects the smallest value of the depth difference
absolute values as the depth difference minimum value for each
pixel. The blur processing section 490 calculates the filter order
from the depth difference minimum value and performs the blur
process in accordance with the calculated filter order. Therefore,
the degree of blur on a subject is changed in accordance with the
depth from the focus area located closest to the subject in the
depth direction. As a result, an image capable of representing a
more natural perspective can be generated even when setting a
plurality of focus areas.
2. Second Embodiment
Configuration Example of Image Capturing Apparatus
[0102] A second embodiment of the present disclosure will be
described with reference to FIG. 12 to FIG. 16. FIG. 12 is a block
diagram illustrating a configuration example of an image capturing
apparatus 101 according to a second embodiment. The image capturing
apparatus 101 is different from the image capturing apparatus 100
of the first embodiment in that the image capturing apparatus 101
generates depth information by the stereo matching method. The
stereo matching method will be described in detail later. The image
capturing apparatus 101 includes a criterion image capturing unit
151 and a reference image capturing unit 152, instead of the image
capturing unit 150. The image capturing apparatus 101 also includes
an image processing device 401 instead of the image processing
device 400. The image capturing apparatus 101 is different from the
image capturing apparatus 100 in that the image capturing apparatus
101 may not include the depth measurement unit 150.
[0103] The criterion image capturing unit 151 captures a criterion
image. The criterion image is an image to be a criterion in
calculating the depth. For example, any one of images including a
right image captured by the right lens of the image capturing
apparatus 101 and a left image captured by the left lens of the
image capturing apparatus 101 may be the criterion image. The
criterion image capturing unit 151 provides the image data storage
unit 200 with criterion image data through the signal line 155. The
criterion image data indicates the captured criterion image.
[0104] The reference image capturing unit 152 captures a reference
image. The reference image is an image which is captured in
synchronization with the criterion image and which is referenced in
calculating the depth. For example, image other than the reference
image of the left and right images may be the reference image. The
reference image capturing unit 152 provides the image data storage
unit 200 with reference image data through the signal line 156. The
reference image data indicates the captured reference image.
[0105] The image processing device 401 reads out the criterion
image data and the reference image data from the image data storage
unit 200. The image processing device 401 calculates the depth from
the criterion and reference image data in relation to the pixel
using the stereo matching method. The image processing device 401
performs the blur process on the criterion image based on the
calculated depth.
[0106] Configuration Example of Image Processing Device
[0107] A configuration example of the image processing device
according to the second embodiment will be described with reference
to FIG. 13. FIG. 13 is a block diagram illustrating an exemplary
configuration of the image processing device 401 according to the
second embodiment. The image processing device 401 is different
from the image processing device 400 of the first embodiment in
that the image processing device 401 includes a depth information
acquisition section 481 instead of the depth information
acquisition section 480.
[0108] FIG. 14 is a block diagram illustrating a configuration
example of the depth information acquisition section 481 according
to the second embodiment. The depth information acquisition section
481 includes a parallax detection part 482 and a depth generation
part 483.
[0109] The parallax detection part 482 detects a parallax from the
criterion and reference images. More specifically, the parallax
detection part 482 reads out the criterion and reference image data
from the image data storage unit 200. The parallax detection part
482 sets any one pixel in the criterion image as a target point,
and obtains a correspondence point corresponding to a target point
in the reference image. The parallax detection part 482 detects an
absolute value of difference between horizontal coordinate of the
target point and horizontal coordinate of the correspondence point
as a parallax. The parallax detection part 482 provides the depth
generation part 483 with the parallax for all pixels in the
criterion image.
[0110] The depth generation part 483 generates depth information
from the parallax. A method for generating the depth information
from the parallax will be described later. The depth generation
part 483 obtains the depth in relation to each pixel in the image
and generates the depth information. The depth generation part 483
provides the blur processing section 490 with the depth
information.
[0111] FIG. 15 is a diagram illustrating an example of a
relationship between parallax and depth according to the second
embodiment. If the horizontal coordinate of the target point in the
criterion image is X.sub.R and the horizontal coordinate of the
correspondence point in the reference image is X.sub.L, the
parallax p is calculated by the following expression:
p=|X.sub.R-X.sub.L| (8)
[0112] If a gap between lenses of the criterion image capturing
unit 151 and the reference image capturing unit 152 is B and focal
length is F, then the triangle formed by the target point
(X.sub.R), the correspondence point (X.sub.L) and a subject is
similar to the triangle formed by the lens of the criterion image
capturing unit 151, the lens of the reference image capturing unit
152, and the subject. As a result, when the depth of the subject is
d, a relational expression which may be indicated as the following
expression is established:
p:F=B:d (9)
[0113] By rearranging Equation (9), the following expression can be
obtained:
d=B.times.F/p (10)
[0114] B and F are known values. By substituting the detected
parallax p into Expression (10), the depth d is calculated.
[0115] Operation Example of Image Processing Device
[0116] An operation example of the image processing device 401
according to the second embodiment will be described with reference
to FIG. 16. FIG. 16 is a flowchart illustrating an operation
example of the image processing device 401 according to the second
embodiment. The operation of the image processing device 401 is
different from the image processing device 400 of the second
embodiment in that the image processing device 401 performs steps
S951 and S952 instead of step S950.
[0117] The image processing device 401 detects parallax p from the
criterion and reference images (step S951). The image processing
device 401 calculates the depth d from the parallax p based on
Expression (10) and generates depth information (step S952).
[0118] Therefore, according to the second embodiment of the present
disclosure, the focus area setting section 410 sets the focus area,
and the depth information acquisition section 481 detects the
parallax from the criterion and reference images and generates the
depth information based on the parallax. The blur processing
section 490 selects the depth difference minimum value for each
pixel based on the depth information and calculates the filter
order from the depth difference minimum value, thereby performing
the blur process. Thus, the image processing device 401 can
independently generate the depth information.
3. Third Embodiment
Configuration Example of Image Processing Device
[0119] A third embodiment of the present disclosure will be
described with reference to FIG. 17 and FIG. 18. FIG. 17 is a block
diagram illustrating a configuration example of a focus area
setting section 411 according to the third embodiment. The focus
area setting section 411 is different from the focus area setting
section 410 of the first embodiment in that the focus area setting
section 411 recognizes a plurality of objects in an image and sets
each of the objects as a focus area. The focus area setting section
411 includes a candidate area extraction part 420, a priority
setting part 450, and a candidate area selection part 460.
[0120] The candidate area extraction part 420 extracts each of the
plurality of areas to be a candidate of focus area as a candidate
area in an image. The candidate area extraction part 420 includes
an object database 421 and an object recognition portion 422.
[0121] The object database 421 is a database for storing
information which is used for recognizing an object to be
recognized. For example, the object database 421 stores various
information such as image data learned about the object to be
recognized, a category of the object, and feature amount of the
object.
[0122] The object recognition portion 422 refers to the object
database 421 and recognizes a plurality of objects in an image. The
object recognition portion 422 sets each area of the recognized
objects to a candidate area which may be a candidate of the focus
area. The object recognition portion 422 generates candidate area
information indicating the candidate area in the image and provides
the candidate area information to the priority setting part 450 and
the candidate area selection part 460.
[0123] The priority setting part 450 sets the priority for
selecting the candidate area to each candidate area. The priority
setting part 450 includes an object information registration
portion 451 and an object priority setting portion 452.
[0124] The object information registration portion 451 registers
object information associated with the object whose priority for
selection is to be changed. More specifically, the object
information may include image data of object, name of object, and
so on.
[0125] The object priority setting portion 452 sets the priority
for each of the candidate areas. For example, the object priority
setting portion 452 sets the priority of the candidate area which
is similar to an image of the registered object to be higher than
that of the candidate area which is not similar to the image of the
registered object. Alternatively, the object priority setting
portion 452 may be configured to set the priority of the candidate
area which is not similar to the image of the registered object to
be higher.
[0126] The object priority setting portion 452 may be configured to
set the priority based on the area size of the candidate area. More
specifically, the object priority setting portion 452 sets the
priority to be higher as the area size of the candidate area
becomes larger. The object priority setting portion 452 may
calculate an average value for the entire area size of the
candidate area and set the priority to be higher as the area size
of the candidate area approaches to the average value.
[0127] The object priority setting portion 452 may set the priority
based on the depth corresponding to the candidate area. In this
case, the object priority setting portion 452 may set the priority
to be higher as the depth is increased. Alternatively, the object
priority setting portion 452 may set the priority to be higher as
the depth is decreased. The object priority setting portion 452 may
obtain a maximum value or a minimum value of the depth
corresponding to the entire pixels in the image and set the
priority of the object to be higher as the depth of the object
approaches the maximum value or the minimum value. In this way,
when a higher priority is assigned to the objects located closer
and farther away in a depth direction, it makes easier to set the
focus area to objects located far away from each other in the depth
direction. As a result, the blurring becomes more and more apparent
as the object deepens. The vicinity of the midpoint of a close
object and a far object is most blurred. The blurring becomes less
and less apparent as the object deepens in the depth direction
beyond the midpoint. This enables the degree of blur to be
noticeable and the perspective in an image is more enhanced.
[0128] The object priority setting portion 452 provides the
candidate area selection part 460 with priority information
indicating the set priority.
[0129] The candidate area selection part 460 selects a plurality of
candidate areas of the candidate areas indicated by the candidate
area information and sets each of the plurality of candidate areas
as a focus area. However, the candidate area selection part 460
limits the number of the focus areas to be set to a value less than
or equal to T (T is an integer greater than or equal to 2). More
specifically, when the number of the candidate areas is limited to
a value less than or equal to T, the candidate area selection part
460 sets all of the candidate areas as the focus area. If the
number of the candidate areas is larger than T, then the candidate
area selection part 460 firstly selects the candidate area which is
set as having a higher priority by the priority setting part 450,
and sets T candidate areas as the focus area. The candidate area
selection part 460 provides the blur processing section 490 with
focus area information indicating the set focus area.
[0130] Although the focus area setting section 411 is configured to
include the object database 421, the focus area setting section 411
may be configured to not include the object database 421. For
example, it is possible that an object database is provided to the
exterior of the image capturing apparatus 100 and the focus area
setting section 411 accesses the object database. In a similar
manner, the object information registration portion 451 may be
provided to the exterior of the focus area setting section 411.
[0131] Although the focus area setting section 411 includes the
priority setting part 450, the focus area setting section 411 may
be configured to not include the priority setting part 450. In this
case, the candidate area selection part 460 may be configured to
set the entire object areas as the focus area without limiting the
number of the focus areas, and the number of the focus areas may be
limited to T. When the priority setting part 450 is not included in
the focus area setting section 411 and the number of the focus
areas is limited to a value less than or equal to T, if the number
of the candidate areas is greater than T, then the candidate area
selection part 460 may select T areas of the object at random.
[0132] Operation Example of Image Processing Device
[0133] An operation example of the image processing device
according to the third embodiment will be described with reference
to FIG. 18. FIG. 18 is a flowchart illustrating an example of a
focus area setting process according to the third embodiment. In
the focus area setting section 411, the object recognition portion
422 recognizes an object in an image with reference to the object
database 421 and sets an area of the object as a candidate area
(step S921). The priority setting part 450 sets a higher priority
to a candidate area similar to an image of the registered object
(step S922). The candidate area selection part 460 determines
whether or not the number of the candidate areas is greater than T
(step S923). If the number of the candidate areas is greater than T
(step S923: Yes), then the candidate area selection part 460
selects T candidate areas based on the priority (step S924).
[0134] If the number of the candidate areas is less than or equal
to T (step S923: No) or after performing step S924, the candidate
area selection part 460 sets each candidate area as a focus area
(step S925). After performing step S925, the focus area setting
section 411 terminates the focus area setting process.
[0135] As described above, according to the third embodiment, the
focus area setting section 411 recognizes an object in an image and
sets an area of the object as a focus area. The depth information
acquisition section 480 generates depth information. The blur
processing section 490 selects a depth difference minimum value for
each pixel based on the depth information and calculates a filter
order from the depth difference minimum value, thereby performing a
blur process. Thus, the image processing device 400 can
independently set the focus area even when a user does not set a
focus area.
4. Fourth Embodiment
Configuration Example of Image Processing Device
[0136] A fourth embodiment of the present disclosure will be
described with reference to FIG. 19 to FIG. 21. FIG. 19 is a block
diagram illustrating a configuration example of a focus area
setting section 412 according to the fourth embodiment. The focus
area setting section 412 differs from the focus area setting
section 410 of the first embodiment in that the focus area setting
section 412 recognizes a plurality of faces in an image and sets
each area of these faces as a focus area. The focus area setting
section 412 includes a candidate area extraction part 430, a
priority setting part 450, and a candidate area selection part
460.
[0137] The candidate area extraction part 430 of the fourth
embodiment differs from the candidate area extraction part 420 of
the third embodiment in that the candidate area extraction part 430
recognizes a face instead of an object. The candidate area
extraction part 430 includes a facial image database 431 and a
facial recognition portion 432.
[0138] The facial image database 431 is a database which stores
information for recognizing a face in an image. The facial image
database 431 stores, for example, facial image data of different
people, data numerically indicating features of facial parts,
metadata such as gender or age of the people, or the like.
[0139] The facial recognition portion 432 recognizes a plurality of
faces in an image with reference to the facial image database 431.
The facial recognition portion 432 sets each area of the recognized
faces as a candidate area.
[0140] The priority setting part 450 of the fourth embodiment
differs from the priority setting part 450 of the third embodiment
in that the priority setting part 450 of the fourth embodiment
gives a higher priority to a candidate area similar to a facial
image of a particular person. The priority setting part 450 of the
fourth embodiment includes a person information registration
portion 453 and a facial area priority setting portion 454.
[0141] The person information registration portion 453 registers
person information relative to a person whose priority is to be
changed in selecting. The person information includes facial image
data of a particular person, metadata such as gender or age of the
person, or the like.
[0142] The facial area priority setting portion 454 sets the
priority to each candidate area. For example, the facial area
priority setting portion 454 recognizes the registered person and
sets the priority of a candidate area similar to a facial image of
the recognized person to be higher than other candidate areas.
[0143] The candidate area selection part 460 of the fourth
embodiment is similar to the candidate area selection part 460 of
the third embodiment.
[0144] The facial image database 431 and the person information
registration portion 453 may be provided to the exterior of the
focus area setting section 412.
[0145] Although the focus area setting section 412 includes the
priority setting part 450, the focus area setting section 412 may
be configured to not include the priority setting part 450.
[0146] Although the priority setting part 450 assigns a higher
priority to a candidate area similar to the facial image of the
registered person, the priority setting part 450 may assign a lower
priority to the candidate area similar to the facial image of the
registered person.
[0147] FIG. 20 is a diagram illustrating an example of image data
before and after blurring according to the fourth embodiment. FIG.
20(a) is a diagram illustrating an example of the image data before
blurring. This image data includes facial areas 631 and 633 and
subjects of trees 632 and 634. The focus area setting section 412
recognizes the facial areas 631 and 633 and sets each facial area
as a focus area. The blur processing section 490 performs the blur
process in accordance with the depth difference minimum value.
[0148] FIG. 20(b) is a diagram illustrating an example of the image
data after blurring. The blur process is not performed on the
subjects 641 and 643 containing facial areas corresponding to the
facial areas 631 and 633. The subject 641 contains other parts of
the body in addition to the face, and the subject 643 contains
other parts of body and a car in addition to the face. Since the
other parts of the bodies and the car are substantially identical
in depth to the faces, the blur process is not performed on the
other parts of the bodies and the car. On the other hand, the blur
process is performed on subjects 642 and 644 corresponding to the
subjects of trees 632 and 634 in accordance with the depth
difference minimum values.
[0149] Operation Example of Image Processing Device
[0150] An operation example of the image processing device 400
according to the fourth embodiment will be described with reference
to FIG. 21. FIG. 21 is a flowchart illustrating an example of a
focus area setting process according to the fourth embodiment. In
the focus area setting section 412, the facial area recognition
portion 432 recognizes a facial area in an image with reference to
the facial image database 431 and sets the facial area as a focus
area (step S931). The priority setting part 450 assigns a higher
priority to a candidate area similar to the facial image of the
registered person (step S932). Step S933 to step S935 are performed
similarly to the step S923 to step S925 in the third
embodiment.
[0151] As described above, according to the fourth embodiment, the
focus area setting section 412 recognizes a facial area in an image
and sets the facial area as a focus area. The depth information
acquisition section 480 generates depth information. The blur
processing section 490 selects a depth difference minimum value for
each pixel based on the depth information and calculates a filter
order from the depth difference minimum value, thereby performing a
blur process. Thus, the image processing device 400 can
independently set the focus area even when a user does not set a
focus area.
5. Fifth Embodiment
Configuration Example of Image Processing Device
[0152] A fifth embodiment of the present disclosure will be
described with reference to FIG. 22 to FIG. 25. FIG. 22 is a block
diagram illustrating a configuration example of a focus area
setting section 413 according to the fifth embodiment. The focus
area setting section 413 differs from the focus area setting
section 410 of the first embodiment in that the focus area setting
section 413 sets the focus area based on a composition.
[0153] A candidate area extraction part 440 of the fifth embodiment
differs from the candidate area extraction part 420 of the third
embodiment in that the candidate area extraction part 440 extracts
a candidate area based on the composition. The candidate area
extraction part 440 includes a composition database 441, a
composition decision portion 442, and a candidate area extraction
processing portion 443.
[0154] The composition database 441 stores information associated
with various compositions such as the rule of thirds composition, a
pattern composition, and a radial composition. For example, the
composition database 441 stores an arrangement pattern in an image
of composition components forming the composition, a name of the
composition, and so on. The composition components include a
subject such as an object or a face, and a line segment detected by
edge detection.
[0155] The composition decision portion 442 decides a composition.
The composition decision portion 442 may detect a subject or a line
segment in an image and select the composition which most closely
matches the shape or arrangement of the subject, thereby deciding a
composition for the image. The composition decision portion 442
provides the candidate area extraction processing portion 443 with
composition information indicating the decided composition. The
composition decision portion 442 may decide the composition in
accordance with the user's operation.
[0156] The candidate area extraction processing portion 443
extracts an area of the main subject in the decided composition as
a candidate area. The candidate area extraction processing portion
443 generates candidate area information indicating the extracted
candidate area and provides the candidate area information to the
priority setting part 450 and the candidate area selection part
460.
[0157] The priority setting part 450 of the fifth embodiment
differs from the priority setting part 450 of the third embodiment
in that the priority setting part 450 of the fifth embodiment sets
the priority of the areas forming various subjects such as an
object or a face. The priority setting part 450 of the fifth
embodiment includes a subject information registration portion 455
and a subject priority setting portion 456.
[0158] The subject information registration portion 455 registers
subject information relative to the subject whose priority is to be
changed in selecting. Specifically, the subject information may be
image data of a particular subject, a name of a subject, or the
like.
[0159] The subject priority setting portion 456 sets the priority
of each candidate area. For example, the subject priority setting
portion 456 sets the priority of candidate area similar to the
image of the recognized subject to be higher than that of the
candidate area which is not similar to the image of the recognized
subject. Alternatively, the subject priority setting portion 456
may set the priority of candidate area similar to the image of the
recognized subject to be lower.
[0160] The candidate area selection part 460 of the fifth
embodiment is similar to the candidate area selection part 460 of
the third embodiment.
[0161] The focus area setting section 413 may be provided to the
exterior of the composition database 441 and the subject
information registration portion 455.
[0162] Although the focus area setting section 413 includes the
priority setting part 450, the focus area setting section 413 may
be configured to not include the priority setting part 450.
[0163] FIG. 23 illustrates examples of image data before and after
blurring when the rule of thirds composition is set, in accordance
with the fifth embodiment. FIG. 23(a) illustrates an example of
image data before blurring 650. In the image data 650, subjects 651
to 654 are captured. FIG. 23(b) illustrates an example of the image
data 650 when the rule of thirds composition is set.
[0164] The rule of thirds composition is a composition in which an
image is equally divided into nine parts by two horizontal lines
and two vertical lines which are drawn at regular intervals and
main subject is placed on the lines or intersection points of the
horizontal and vertical lines. By employing the rule of thirds
composition, it is possible to attain a well-balanced artistic
image.
[0165] It is assumed that the composition decision portion 442
selects any one of the rule of thirds composition and the pattern
composition. It is assumed that the subjects 651 to 654 are
different in shape. In this case, the rule of thirds composition is
selected because the pattern composition is not suitable. In the
rule of thirds composition, intersection points PP1 to PP4 are set.
As shown in FIG. 23(b), the subjects 651 and 654 are both not
located at any intersection point. On the other hand, the subject
652 is located at the intersection point PP2 and the subject 653 is
located at the intersection point PP3. Thus, the subjects 652 and
653 are determined as main subjects for the composition and each
subject is set as a focus area.
[0166] FIG. 23(c) illustrates an example of an image data after
blurring 660. Subjects 661 to 664 in the image data 660 correspond
to the subjects 651 to 654 in the image data before blurring 650,
respectively. Because each of the subjects 652 and 653 is set as a
focus area, the blur process is not performed on subjects 662 and
663 respectively corresponding to these subjects 652 and 653. On
the other hand, the blur process is performed on subjects 661 and
664 in accordance with the depth difference minimum value.
[0167] FIG. 24 illustrates examples of focus area information and
image data before and after blurring when the pattern composition
is set, in accordance with the fifth embodiment. FIG. 24(a)
illustrates an example of image data before blurring 670. Subjects
671 to 674 are captured in the image data 670.
[0168] The pattern composition is a composition for setting the
subjects which are regularly arranged in an image and similar to
each other in shape as a main subject. It is possible to attain a
rhythmic image by employing the pattern composition. For example,
in FIG. 24(a), subjects 671 to 674 are arranged regularly. The
subjects 671, 672 and 674 have a similar shape. In this case,
because there are the subjects 671, 672 and 674 which are regularly
arranged and similar to each other in shape, the pattern
composition is selected. Thus, the subjects 671, 672 and 674 are
determined as main subjects for the composition and each subject is
set as a focus area.
[0169] FIG. 24(b) illustrates focus area information 680. In FIG.
24(b), the pixel value of the white area is zero (0) and the pixel
value of the gray area is one (1). In the focus area information
680, focus areas 681, 682 and 684 respectively corresponds to the
subjects 671, 672 and 674 which each are set as the focus area in
the image data 670. As shown in FIG. 24(b), the pixel value of
pixel in the focus area is set to zero (0), the pixel value of
pixel of other area is set to one (1).
[0170] FIG. 24(c) illustrates an image data after blurring 690.
Subjects 691 to 694 in image data 690 respectively corresponds to
the subjects 671 to 674 in the image data 670. Because each of the
subjects 671, 672 and 674 is set as a focus area, the blur process
is not performed on subjects 691, 692 and 694 respectively
corresponding to the subjects 671, 672 and 674. On the other hand,
the subject 693 is not set as the focus area, thus the blur process
is performed on the subject in accordance with the depth difference
minimum value.
[0171] Operation Example of Image Processing Device
[0172] An operation example of the image processing device
according to the fifth embodiment will be described with reference
to FIG. 25. FIG. 25 is a flowchart illustrating an example of a
focus area setting process according to the fifth embodiment. In
the focus area setting section 413, the composition decision
portion 442 decides a composition based on the shape or arrangement
of the subject in an image (step S941). The candidate area
extraction processing portion 443 sets each of a plurality of
subject areas as a candidate area based on the composition (step
S942). The priority setting part 450 sets a higher priority to the
candidate area similar to the image of the registered subject (step
S943). Step S944 to step S946 are performed similarly to the step
S923 to step S925 in the third embodiment.
[0173] As described above, according to the fifth embodiment, the
focus area setting section 413 decides a composition and sets the
focus area based on the composition. The depth information
acquisition section 480 generates depth information. The blur
processing section 490 selects a depth difference minimum value for
each pixel based on the depth information and calculates a filter
order from the depth difference minimum value, thereby performing a
blur process. Thus, the image processing device 400 can
independently set the focus area even when a user does not set a
focus area.
[0174] It should be understood by those skilled in the art that
various modifications, combinations, sub-combinations and
alterations may occur depending on design requirements and other
factors insofar as they are within the scope of the appended claims
or the equivalents thereof.
[0175] In addition, the process procedures in the above-described
embodiments may be grasped as a method including the series of
procedures, and may be grasped as a program for enabling a computer
to execute the series of procedures or a recording medium storing
the program thereon. The recording medium may use, for example, a
CD (Compact Disc), an MD (Mini Disc), a DVD (Digital Versatile
Disk), a memory card, a Blu-ray Disc (registered trademark), or the
like.
[0176] Additionally, the present technology may also be configured
as below.
(1) An image processing device including:
[0177] a focus area setting section for setting each of a plurality
of areas that are to be kept in focus in a captured image of a
subject as a focus area;
[0178] a depth acquisition section for acquiring a depth of the
subject in relation to each pixel in the image; and
[0179] a blur processing section for setting each pixel in the
image as a target pixel and for performing a blur process on the
target pixel in accordance with a minimum value of each depth
difference absolute value between depth corresponding to the target
pixel and depth corresponding to each of the focus area.
(2) The image processing device according to (1),
[0180] wherein the focus area setting section includes
[0181] a candidate area extraction part for extracting each of a
plurality of areas to be a candidate for the focus area in the
image as a candidate area, and
[0182] a candidate area selection part for selecting a plurality of
candidate areas among the extracted candidate areas and for setting
each of the plurality of candidate areas as the focus area.
(3) The image processing device according to (2), wherein the
candidate area extraction part recognizes a captured area of a
given object in the image and extracts the area as the candidate
area. (4) The image processing device according to (2) or (3),
wherein the candidate area extraction part recognizes a captured
area of a face in the image and extracts the area as the candidate
area. (5) The image processing device according to any of (2) to
(4),
[0183] wherein the candidate area extraction part includes
[0184] a composition decision portion for deciding a composition to
define an arrangement of the subject in the image, and
[0185] a candidate area extraction processing portion for
extracting a plurality of areas as the candidate area based on the
decided composition.
(6) The image processing device according to any of (2) to (5),
[0186] wherein the focus area setting section further includes
[0187] a priority setting part for setting a priority for the
selection of the candidate area to the extracted candidate area,
and
[0188] wherein the candidate area selection part selects the
candidate area based on the set priority.
(7) The image processing device according to (6), wherein the
priority setting part registers a captured image of a given subject
and sets the priority to the candidate area based on a similarity
between the registered image and the candidate area. (8) The image
processing device according to (6) or (7), wherein the priority
setting part sets the priority based on an area size of the
candidate area. (9) The image processing device according to any of
(6) to (9), wherein the priority setting part sets the priority
based on the depth corresponding to any one pixel in the candidate
area. (10) The image processing device according to (9), wherein
the depth setting part sets the priority of the candidate area to
be higher as the depth corresponding to any one pixel in the
candidate area is closer to a maximum value or minimum value of the
obtained depth. (11) The image processing device according to any
of (1) to (10),
[0189] wherein the image includes a criterion image and a reference
image for using in the acquisition of the depth,
[0190] wherein the depth acquisition section includes
[0191] a parallax detection part for detecting a distance between
any one pixel in the criterion image and a pixel, corresponding to
the pixel in the criterion image, in the reference image as a
parallax, and
[0192] a depth generation part for generating the depth in relation
to the pixel in the criterion image based on the detected parallax,
and
[0193] wherein the blur processing section performs the blur
process on each of the pixels in the criterion image as the target
pixel.
(12) The image processing device according to any of (1) to (11),
wherein the blur processing section performs the blur process on
the target pixel, the blur process being calculated by using the
following expression:
.delta..sub.j=D.times.d.sub.j--min.times.R/d.sub.j
where .delta..sub.j is a degree of blur, D is an effective diameter
of an image capturing apparatus by which the image is captured,
d.sub.j is the depth corresponding to the target pixel,
.DELTA.d.sub.j--min is the minimum value, and R is an imaging
magnification of the image. (13) The image processing device
according to any of (1) to (12), wherein the focus area setting
section sets the focus area in accordance with an operation signal
for specifying the focus area. (14) The image processing device
according to any of (1) to (13), wherein the focus area setting
section sets the focus area by limiting the number of focus areas
to be set to a value less than or equal to a threshold. (15) A
method of controlling an image processing device, including:
[0194] setting, by a focus area setting section, each of a
plurality of areas that are to be kept in focus in a captured image
of a subject as a focus area;
[0195] acquiring, by a depth acquisition section, a depth of the
subject in relation to each pixel in the image; and
[0196] setting, by a blur processing section, each pixel in the
image as a target pixel and performing a blur process on the target
pixel in accordance with a minimum value of each depth difference
absolute value between depth corresponding to the target pixel and
depth corresponding to each of the focus area.
(16) A program for enabling a computer to execute a procedure, the
procedure including:
[0197] setting, by a focus area setting section, each of a
plurality of areas that are to be kept in focus in a captured image
of a subject as a focus area;
[0198] acquiring, by a depth acquisition section, a depth of the
subject in relation to each pixel in the image; and
[0199] setting, by a blur processing section, each pixel in the
image as a target pixel and performing a blur process on the target
pixel in accordance with a minimum value of each depth difference
absolute value between depth corresponding to the target pixel and
depth corresponding to each of the focus area.
[0200] The present disclosure contains subject matter related to
that disclosed in Japanese Priority Patent Application JP
2011-142807 filed in the Japan Patent Office on Jun. 28, 2011, the
entire content of which is hereby incorporated by reference.
* * * * *