U.S. patent application number 09/782253 was filed with the patent office on 2006-09-14 for image processing apparatus for correcting contrast of image.
Invention is credited to Nobuhiro Aihara.
Application Number | 20060204124 09/782253 |
Document ID | / |
Family ID | 18573100 |
Filed Date | 2006-09-14 |
United States Patent
Application |
20060204124 |
Kind Code |
A1 |
Aihara; Nobuhiro |
September 14, 2006 |
IMAGE PROCESSING APPARATUS FOR CORRECTING CONTRAST OF IMAGE
Abstract
In a contrast correcting apparatus 1 which divides an image into
unit regions and carries out a contrast correction for each of unit
regions, a gray level histogram calculation section 201 generates
gray level histograms of the image, and a scene judgment section
202 makes a judgment on the state of the image. When the result of
the judgment shows that the image is in a state such as
overexposure, underexposure, low contrast or high contrast, a
region size determining section 204 sets a greater size as the size
of the unit regions. Based upon the size of the unit regions thus
determined and the amount of contrast correction determined by the
contrast correction amount determining section 203, gray level
transformation curves are formed for the respective unit regions,
and by using these, the contrast corrections are carried out on the
respective unit regions. With this method, it is possible to
properly reduce the occurrence of unevenness in gray levels in an
image that tends to arise in the case when the size of the unit
regions is small although the amount of contrast correction (the
amount of emphasis) is great.
Inventors: |
Aihara; Nobuhiro;
(Amagasaki-Shi, JP) |
Correspondence
Address: |
BUCHANAN, INGERSOLL & ROONEY PC
POST OFFICE BOX 1404
ALEXANDRIA
VA
22313-1404
US
|
Family ID: |
18573100 |
Appl. No.: |
09/782253 |
Filed: |
February 14, 2001 |
Current U.S.
Class: |
382/274 ;
345/617; 348/254; 348/E5.074; 382/167 |
Current CPC
Class: |
H04N 1/40062 20130101;
H04N 1/4074 20130101; G06T 5/008 20130101; G06T 2207/20021
20130101; H04N 5/202 20130101; G06T 5/40 20130101 |
Class at
Publication: |
382/274 ;
382/167; 345/617; 348/254 |
International
Class: |
G06K 9/00 20060101
G06K009/00; H04N 5/202 20060101 H04N005/202; G06K 9/40 20060101
G06K009/40; G09G 5/00 20060101 G09G005/00 |
Foreign Application Data
Date |
Code |
Application Number |
Feb 28, 2000 |
JP |
2000-051446 |
Claims
1. A computer-readable medium carrying a program for correcting a
contrast of an image, wherein execution of said program by a
computer causes said computer to perform a process comprising the
steps of: determining a size of unit regions each of which is a
unit for correcting said contrast of said image; dividing said
image into a plurality of unit regions in accordance with said size
of unit regions; obtaining a plurality of gray level transformation
characteristics, each gray level transformation characteristic
individually corresponding to a different one of said plurality of
unit regions; and correcting the contrast of each said unit region
by individually using said gray level transformation characteristic
individually corresponding to said unit region.
2. The computer-readable medium of claim 1, wherein said size of
unit regions is determined on the basis of distribution of a gray
level histogram generated from said image.
3. The computer-readable medium of claim 1, wherein said size of
unit regions is determined on the basis of pickup conditions at the
time of picking up said image.
4. The computer-readable medium of claim 1, wherein said size of
unit regions is determined on the basis of information input by an
operator.
5. The computer-readable medium of claim 1, wherein said size of
unit regions is increased as an amount of contrast correction of
said image becomes greater.
6. The computer-readable medium of claim 1, wherein said size of
unit regions is determined on the basis of a size of a region in
which values of a predetermined color element are located within a
predetermined range.
7. A method of correcting a contrast of an image comprising the
steps of: determining a size of unit regions each of which is a
unit for correcting said contrast of said image; dividing said
image into a plurality of unit regions in accordance with said size
of unit regions; obtaining a plurality of gray level transformation
characteristics, each gray level transformation characteristic
individually corresponding to a different one of said plurality of
unit regions; and correcting contrasts of each said of unit region
by individually using said gray level transformation characteristic
individually corresponding to said unit region.
8. The method of claim 7, wherein said size of unit regions is
determined on the basis of distribution of a gray level histogram
generated from said image.
9. The method of claim 7, wherein said size of unit regions is
determined on the basis of pickup conditions at the time of picking
up said image.
10. The method of claim 7, wherein said size of unit regions is
determined on the basis of information input by an operator.
11. The method of claim 7, wherein said size of unit regions is
increased as an amount of contrast correction of said image becomes
greater.
12. The method of claim 7, wherein said size of unit regions is
determined on the basis of a size of a region in which values of a
predetermined color element are located within a predetermined
range.
13. An apparatus for correcting a contrast of an image comprising:
means for determining a size of unit regions each of which is a
unit for correcting said contrast of said image; means for dividing
said image into a plurality of unit regions in accordance with said
size of unit regions and obtaining a plurality of gray level
transformation characteristics, each gray level transformation
characteristic individually corresponding to a different one of
said plurality of unit regions; and means for correcting contrasts
of each said of unit region by individually using said individually
corresponding gray level transformation characteristic.
14. The apparatus of claim 13, wherein said size of unit regions is
determined on the basis of distribution of a gray level histogram
generated from said image.
15. The apparatus of claim 13, wherein said size of unit regions is
determined on the basis of pickup conditions at the time of picking
up said image.
16. The apparatus of claim 13, wherein said size of unit regions is
determined on the basis of information input by an operator.
17. The apparatus of claim 13, wherein said size of unit regions is
increased as an amount of contrast correction of said image becomes
greater.
18. The apparatus of claim 13, wherein said size of unit regions is
determined on the basis of a size of a region in which values of a
predetermined color element are located within a predetermined
range.
19. An apparatus for correcting a contrast of an image comprising:
a region size determining section for determining a size of unit
regions each of which is a unit for correcting said contrast of
said image; a gray level transformation section for dividing said
image into a plurality of unit regions in accordance with said size
of unit regions and obtaining a plurality of gray level
transformation, each gray level transformation characteristic
individually corresponding to a different one of said plurality of
unit regions; and a contrast correction section for correcting
contrasts of each of said unit regions by individually using said
individually corresponding gray level transformation
characteristics.
Description
[0001] This application is based on application No. 2000-051446
filed in Japan, the contents of which are hereby incorporated by
reference.
BACKGROUND OF THE INVENTION
[0002] 1. Field of the Invention
[0003] The present invention relates to a technique for correcting
contrast of images.
[0004] 2. Description of the Background Art
[0005] Conventionally, a contrast correcting operation for
emphasizing contrast of images picked up by a digital camera or a
scanner has been carried out. Here, in general, the contrast refers
to a difference in gray levels between a background part and a main
subject part in an image or a difference between the maximum gray
level and the minimum gray level in an image; in this case,
however, it is assumed that the contrast simply refers to the
degree of differences between bright portions and dark portions
distributed in an image. Therefore, in the following description,
the correction of the contrast substantially corresponds to a
process for correcting a gray level histogram of an image for
transforming the gray levels in pixels.
[0006] For one example of contrast corrections, a method has been
proposed, in which a gray level histogram is generated in a
predetermined region around a pixel of interest that is to be
corrected, and after the gray level histogram has been transformed,
an accumulated histogram is generated, and the resulting
accumulated histogram is utilized as a gray level transformation
curve so as to transform the gray level of the pixel of
interest.
[0007] However, in this method (hereinafter, referred to as
"correction method with local histograms), it is necessary to
generate gray level histograms for respective pixels, and the
resulting problem is that a great amount of calculations have to be
made.
[0008] As to a modified method for the correction method with local
histograms, another method has been proposed, in which an image is
divided into a plurality of rectangular regions, and a gray level
histogram is formed for each region, and this gray level histogram
is transformed to an accumulated histogram, and by utilizing this
as a gray level transformation curve, the gray levels of all the
pixels within each region are transformed. With this method, the
gray level transformation curves as many as the number of the
divided regions are generated to correct the contrast, thereby
making it possible to shorten the calculation time.
[0009] However, in the conventional contrast correction method
which divides an image into a plurality of regions, and carries out
a contrast correcting process on the basis of each region, the size
of the region is fixed. As a result, in the case when gray level
transformation curves are greatly different between the regions,
the gray levels are greatly different on both of the sides of the
border between the corresponding regions. Here, it is of course
possible to alleviate the difference in gray levels between the
regions by using linear interpolation, etc. however, in the case
when gray levels are greatly different on both of the sides of the
border, unevenness in gray levels become conspicuous even in an
image after the correction.
SUMMARY OF THE INVENTION
[0010] The object of the present invention is to reduce unevenness
in gray levels appropriately upon correcting the contrast of an
image.
[0011] The present invention is directed to a computer-readable
medium carring a program for correcting a contrast of an image.
[0012] According to an aspect of the present invention, execution
of the program by a computer causes the computer to perform a
process comprising the steps of: determining a size of unit regions
each of which is a unit for correcting the contrast of the image;
dividing the image into a plurality of unit regions in accordance
with the size of unit regions; obtaining a plurality of gray level
transformation characteristics corresponding to the plurality of
unit regions, respectively; and correcting contrasts of the
plurality of unit regions by using the plurality of gray level
transformation characteristics.
[0013] The contrast is corrected appropriately and unevenness in
gray levels is reduced by determining the size of unit regions
properly.
[0014] Preferably, the size of unit regions is determined on the
basis of distribution of a gray level histogram generated from the
image, pickup conditions at the time of picking up the image,
information input by an operator, an amount of contrast correction
of the image, or a size of a region in which values of a
predetermined color element are located within a predetermined
range.
[0015] The present invention is also directed to a method and
apparatus for correcting a contrast of an image.
[0016] These and other objects, features, aspects and advantages of
the present invention will become more apparent from the following
detailed description of the present invention when taken in
conjunction with the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0017] FIG. 1 is a drawing that shows a computer which serves as a
contrast correcting apparatus, and its peripheral apparatuses;
[0018] FIG. 2 is a block diagram that shows an inner construction
of the computer;
[0019] FIG. 3 is a block diagram that shows functional components
of the contrast correcting apparatus in the first preferred
embodiment;
[0020] FIGS. 4 and 5 are flowcharts that show a sequence of
processes in the contrast correcting apparatus in the first
preferred embodiment;
[0021] FIGS. 6 and 7 are drawings that show gray level
histograms;
[0022] FIG. 8 is a drawing that shows unit regions;
[0023] FIGS. 9 and 10 are drawings that explain states in which a
gray level transformation curve is formed;
[0024] FIG. 11 is a drawing that explains the gray level
transformation curve;
[0025] FIG. 12 is a drawing that explains a state in which the gray
level transformation curve is interpolated;
[0026] FIGS. 13 and 14 are drawings that explain states in which a
gray level transformation curve is formed;
[0027] FIG. 15 is a drawing that explains the gray level
transformation curve;
[0028] FIG. 16 is a drawing that shows unit regions with a smaller
size;
[0029] FIG. 17 is a drawing that shows a gray level histogram in a
unit region;
[0030] FIG. 18 is a drawing that shows a gray level transformation
curve in the unit region;
[0031] FIG. 19 is a drawing that shows a gray level histogram in
another unit region;
[0032] FIG. 20 is a drawing that shows a gray level transformation
curve in another unit region;
[0033] FIG. 21 is a drawing that shows unit regions with a larger
size;
[0034] FIG. 22 is a block diagram that shows one portion of
functional components of a contrast correcting apparatus in
accordance with the second preferred embodiment;
[0035] FIG. 23 is a drawing that shows one portion of the operation
of the contrast correcting apparatus in accordance with the second
preferred embodiment;
[0036] FIG. 24 is a block diagram that shows one portion of
functional components of a contrast correcting apparatus in
accordance with the third preferred embodiment; and
[0037] FIG. 25 is a drawing that shows one portion of the operation
of the contrast correcting apparatus in accordance with the third
preferred embodiment.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
1. First Preferred Embodiment
[0038] FIG. 1 is a drawing that shows a contrast correcting
apparatus 1 and its peripheral apparatuses in accordance with the
first preferred embodiment. The contrast correcting apparatus 1 is
an apparatus for correcting the gray levels of respective pixels in
an image so as to provide proper contrast corresponding to the
degree of differences between bright portions and dark portions in
the image.
[0039] The contrast correcting apparatus 1 shown in FIG. 1 is
realized mainly by a computer 10, and to the computer 10 are
connected a key board 11a and a mouse 11b through which inputs from
the operator are accepted. Moreover, a display 91 and a printer 92
are connected to the computer 10, and image data can be input from
a digital camera 93 thereto through a memory card, a communication,
cable etc.
[0040] In order to allow the computer 10, the key board 11a and the
mouse 11b to function as the contrast correcting apparatus 1, a
program for correcting contrast is preliminarily installed into the
computer 10 through a recording medium 8, such as an optical disk,
a magnetic disk, a magneto-optical disk and a memory card. Here,
the program may be installed through a computer communication such
as the Internet. In this case, a contrast correction program inside
a recording device of a sender is transferred into the computer 10
through a Web server.
[0041] FIG. 2 is a block diagram that shows an inner construction
of the computer 10 together with its peripheral devices. As
illustrated in FIG. 2, the computer 10, which has the same
construction as a normal computer, is provided with a CPU 101 for
executing various calculations, a ROM 102 for storing a basic
program, a RAM 103 for storing a contrast correction program 131
and for providing a work area of the calculations, etc. that are
connected to a bus line. Moreover, to the bus line are connected
through an interface (I/F) on demand a display 91 and a printer 92
that are peripheral apparatuses, a fixed disk 104 for storing
various programs including the contrast correction program 131, a
readout section 105 for reading programs, etc. from the recording
medium 8, a card slot 106 for transmitting and receiving image data
to and from the digital camera 93 through a memory card, and a key
board 11a and a mouse 11b for accepting inputs from the
operator.
[0042] The contrast correction program 131 is taken into the fixed
disk 104 through the readout section 105 (when obtained through
communication, through a communication section placed separately),
and the program 131 is copied to the RAM 103. Thus, the CPU 101
executes calculation processes in accordance with the program 131
so that the computer 10 to which the key board 11a and the mouse
11b are connected is allowed to function as the contrast correcting
apparatus 1.
[0043] FIG. 3 is a block diagram that shows functional components
realized by the CPU 101, ROM 102, RAM 103, etc. shown in FIG. 2,
together with peripheral devices. In FIG. 3, data which are
transmitted and received are also illustrated on demand. Moreover,
FIGS. 4 and 5 are flowcharts that show a sequence of processes of
the contrast correcting apparatus 1. Referring to FIGS. 3 to 5, an
explanation will be given of the operation of the contrast
correcting apparatus 1.
[0044] First, the contrast correcting apparatus 1 inputs image data
51 from the digital camera 93 to the RAM 103 (step S11). In this
case, an image obtaining device other than the digital camera 93,
such as a scanner, may of course be used, and image data 51,
preliminarily stored in the fixed disk 104, may be transferred into
the RAM 103.
[0045] Upon completion of the preparation of the image data 51, a
gray level histogram calculation section 201 transforms pixel
values of an image indicated by the image data 51 from RGB values
to HSL values (including factors such as the hue, saturation,
lightness (gray level)) (step S12). Then, a histogram (hereinafter,
referred to as a "gray level histogram") of the number of pixels
corresponding to the gray levels of the entire image is generated
(step S13). Here, although an actual object to be processed by the
CPU 101 is "image data", it is referred to simply as "image" in the
following explanation.
[0046] When the gray level histogram has been generated, a scene
judgment section 202 makes a judgment as to a state of the image
based upon the gray level histogram (step S14). The states of an
image include underexposure, overexposure, low contrast, high
contrast, etc., and in the following description, information
indicating any of these image states is referred to as scene
information. A plurality of pieces of scene information are
preliminarily prepared, and the scene judgment section 202 properly
makes a selection on the pieces of scene information based upon the
gray level histogram.
[0047] FIGS. 6 and 7 are drawings that show examples of the gray
level histograms obtained from the image. In these figures, the
minimum value of the gray levels is set to 0, and the maximum value
thereof is set to 255. In general, in the case of an image having
an appropriate contrast, with respect to gray level values, the
gray level histogram exhibits a virtually constant state (or, a
state in which the central portion is slightly raised). However, as
is indicated by symbol 51a in FIG. 6, in the case when the
histogram distribution is biased onto the low gray level region,
the scene judgment section 202 makes a judgment that the image in
question is in an underexposure state, and as is indicated by
symbol 52a, in the case when the histogram distribution is biased
onto the high gray level region, it makes a judgment that the image
in question is in an overexposure state.
[0048] Moreover, as is indicated by symbol 53a, in the case when
the dispersion of the histograms is small with a distribution
biased onto the intermediate gray level region, the scene judgment
section 202 makes a judgment that the image in question is in a low
contrast state, and as is indicated by symbol 54a in FIG. 7, in the
case when the histogram is distributed onto the high gray level
region and low gray level region in a separated manner, it makes a
judgment that the image is in a high contrast state.
[0049] The scene information selected by the scene judgment section
202 is sent to a contrast correction amount determining section 203
in which the amount of contrast correction is determined (step
S15). The amount of contrast correction refers to a parameter
indicating the degree of the contrast correction in an image; and
in this case, for convenience of explanation, it is assumed that
the same contrast correction amount is used for the entire image.
Here, the image data 51 may also be input to the contrast
correction amount determining section 203, and not only the scene
information, but also the image data 51 may be used for determining
the contrast correction amount.
[0050] The scene information is also sent to an region size
determining section 204 where the size of divided regions to be
used at the time of dividing an image is determined (step S16).
[0051] FIG. 8 is a drawing that shows an example for dividing an
image, and each of divided rectangular regions 61 forms a unit
based on which the gray level is transformed by using the same gray
level transformation characteristic (gray level transformation
curve, which will be described later). In the following
description, these regions, each forming the unit based on which
the contrast correction is carried out, are referred to as "unit
regions".
[0052] Thereafter, the size of the unit regions 61 thus determined
is sent to a gray level transformation curve forming section 205
where the image is divided in accordance with the size of the unit
regions 61 (step S117).
[0053] Next, one unit region 61 is specified as a region of
interest, and a gray level histogram of the region of interest is
generated by the gray level transformation curve forming section
205 (step S21). FIG. 9 is a drawing that shows an example of a gray
level histogram of the region of interest. A clip value shown in
FIG. 9 corresponds to the amount of contrast correction that is
given from the contrast correction amount determining section 203
to the gray level transformation curve forming section 205, and the
gray level transformation curve forming section 205 transforms the
gray level histogram 501 by using the clip value.
[0054] FIG. 10 is a drawing that shows the gray level histogram 502
that have been transformed. In FIG. 10, a region 512 indicated by
parallel slanting lines has the same area as a region 511 that
exceeds the clip value in FIG. 9. In other words, a gray level
histogram 502 shown in FIG. 10 is formed by eliminating the region
511 from the gray level histogram 501 of FIG. 9 and adding the
region 512 thereto. The operation for properly clipping the gray
level histogram by a clip value is carried out in order to reduce
an excessive contrast emphasis in the case when an accumulated
histogram, which will be described later, is utilized as the gray
level transformation curve.
[0055] Successively, as illustrated in FIG. 11, an accumulated
histogram 701 is formed with respect to the gray level histogram
502 by the gray level transformation curve forming section 205
(step S22). Then, in FIG. 11, input gray levels from 0 to 255 are
plotted on the axis of abscissa while output gray levels from 0 to
255 are plotted on the axis of ordinate so that the accumulated
histogram 701 is utilized as the gray level transformation curve
for transforming the gray levels of the respective pixels within
the region of interest. Here, in fact, the gray level
transformation curve is formed as a transformation table.
[0056] When the gray level transformation curve has been formed
with respect to the region of interest, the region of interest is
switched to the next unit region 61, and calculations for the gray
level transformation curve are again carried out. Thereafter, the
switchover is successively made with respect to the regions of
interest so that the gray level transformation curves are formed
with respect to all the unit regions 61 (step S23).
[0057] When a plurality of gray level transformation curves have
been formed, the gray levels of the respective pixels are
transformed by using the corresponding gray level transformation
curves. More specifically, one pixel of interest is determined, and
the gray level of the pixel of interest is transformed by using the
gray level transformation curve of the unit region 61 to which the
pixel of interest belongs by a contrast correcting section 206
(step S24). Then, the switchover is successively made with respect
to the pixels of interest so that the transformations of gray
levels are carried out on the entire image, that is, the correction
of contrast is carried out (step S25).
[0058] Here, since the gray level transformation curves are
different according to the respective unit regions 61, an
interpolation process may be carried out on the gray level
transformation curve upon transformation of the gray levels so that
the differences in gray levels may not become conspicuous on both
sides of the borders of the unit regions 61.
[0059] FIG. 12 is a drawing that explains one example for the
interpolation process of the gray level transformation curve. In
FIG. 12, on the assumption that, upon transforming the gray level
of a pixel 641 in a unit region 604, the gray level transformation
curves of adjacent unit regions 601, 602, 603 and 604 are
represented by functions f(x), g(x), h(x), i(x) (x: input gray
level), and that distances from the centers of gravity of the
respective unit regions 601, 602, 603 and 604 to the pixel 641
(distances indicated by symbols 611, 612, 613 and 614) are
represented by a, b, c and d, the gray level transformation curve
after the interpolation is found from the following expression 1: d
f .times. .times. ( x ) + c g .times. .times. ( x ) + b h .times.
.times. ( x ) + a i .times. .times. ( x ) a + b + c + d [
Expression .times. .times. 1 ] ##EQU1##
[0060] Such an interpolation process is carried out at the time of
the gray level transformation so that the differences in gray
levels on both sides of the borders of the unit regions can be
alleviated. Here, with respect to the interpolation process,
various other methods may of course be utilized.
[0061] Upon completion of the contrast correction on the image,
pixel values of the respective pixels are transformed from the HSL
values to values having a format suitable for the output apparatus.
For example, in the case when the image is displayed on the display
91, the HSL values are transformed to RGB values, and in the case
when it is printed by the printer 92, they are transformed to CMYK
values (step S26). Thereafter, the image data 52 representing the
image that has been corrected is output to a specified output
apparatus (step S27).
[0062] In the above description, an explanation has been given of
the operation of the contrast correcting apparatus 1. Next, an
explanation will be given of the functions of the contrast
correction amount determining section 203 and the region size
determining section 204 shown in FIG. 3.
[0063] FIGS. 13 through 15 are drawings that correspond to FIGS. 9
through 11, and FIGS. 13 and 14 are drawings that show states of a
transformation of a gray level histogram in the case of a clip
value greater than that of FIG. 9. In other words, in the gray
level histogram 503 in FIG. 13, the area of a region 513 exceeding
a clip value is set to be equal to the area of a region 514 of the
gray level histogram 504 in FIG. 14.
[0064] In the case when the clip value is great, the area of the
region 513 becomes smaller than the region 511 shown in FIG. 9, and
as illustrated in FIG. 15, the accumulated histogram 702 of the
gray level histogram 504 forms a curve that has an abrupt maximum
inclination as shown in FIG. 15. Therefore, when the gray level
transformation curve shown in FIG. 15 is used to carry out a gray
level transformation, the degree of contrast emphasis becomes
higher as compared with a case in which the gray level
transformation curve shown in FIG. 11 is used to carry out a gray
level transformation. In this manner, in general, the greater the
clip value, the higher the degree of contrast emphasis.
[0065] Moreover, as described earlier, the clip value (that is, the
amount of contrast corrections is determined by the contrast
correction amount determining section 203 based upon the scene
information and the image data. The scene information includes
underexposure, overexposure, low contrast, high contrast, etc., and
in general, in the case of an image from which any piece of scene
information is obtained, since there is a bias on the distribution
of the gray levels, the image is subjected to low contrast
partially or entirely. Therefore, in the case when any of those
pieces of scene information is obtained, an amount of contrast
correction which is greater than that of a normal image is
determined by the contrast correction amount determining section
203.
[0066] In the case of scene information for which an amount of
contrast correction greater than that of the normal image has to be
determined, the region size determining section 204 determines the
size of the unit region so as to be greater than that of the normal
image.
[0067] FIGS. 16 through 20 explain problems that arise in the case
when, although the amount of contrast correction has been
increased, the size of the unit region is maintained as it is. In
the image shown in FIG. 16, the unit region 62 is set as a region
including only the background, and when the gray level histogram is
generated, the resulting histogram has a biased portion on the
highlighted side as illustrated in FIG. 17. Therefore, the
resulting gray level transformation curve is given as a curve
having a maximum inclination on the highlighted side approximately
shown in FIG. 18.
[0068] In contrast, the unit region 63 is set as a region including
the background and a main subject part, and when the gray level
histogram is generated, the resulting histogram has a plurality of
peaks as shown in FIG. 19. Therefore, the resulting gray level
transformation curve is given as a curve approximately shown in
FIG. 20, which is greatly different from that of FIG. 18.
[0069] As illustrated in FIG. 21, in the case when the size of the
unit region is increased, since the background and the main subject
part are contained in both of a unit region 64 and a unit region 65
in FIG. 21, the difference in the gray level transformation curves
derived from these unit regions becomes smaller. In other words, as
the unit region is increased, the gray level transformation curve
is derived from image information of a wider range so that the
difference in the gray level transformation curves of the
respective unit regions is reduced.
[0070] As described above, the difference in the gray level
transformation curves (that is, transformation characteristics)
between the unit regions becomes more conspicuous as the size of
the unit region becomes smaller. Moreover, in general, the
difference becomes more conspicuous as the amount of contrast
correction (clip value) becomes greater. Therefore, in the case
when, although the amount of the contrast correction is increased
so as to carry out a contrast correction to which even detailed
parts of an image are reflected, the size of the unit regions is
still maintained at the same size as a normal image, a problem
arises that gray levels are greatly different on both sides of the
borders among the unit regions (that is, unevenness in gray levels
in the image becomes conspicuous).
[0071] Therefore, in the contrast correcting apparatus 1 in
accordance with the present preferred embodiment, in the case when
any scene information that makes the amount of contrast correction
greater is obtained, the region size determining section 204
determines a greater size as the size of the unit regions. As a
result, it becomes possible to properly reduce unevenness in the
gray levels.
[0072] For example, in the case when, in a normal amount of
contrast correction, a size of pixels of 64.times.64 is used as the
size of the unit regions, and in the case of a greater amount of
contrast correction, the size of the unit regions is set to a size
of pixels of 128.times.128. In the case of a further greater amount
of contrast correction, the size of the unit regions is set to a
size of pixels of 256.times.256. In this manner, the size of the
unit regions is increased as the amount of contrast correction
becomes greater.
2. Second Preferred Embodiment
[0073] In the first preferred embodiment, the scene information is
determined based upon the distribution of the gray level histogram
of an image; and in the second preferred embodiment, an explanation
will be given of a case in which another scene information is
utilized.
[0074] FIG. 22 is a block diagram that shows one portion of
functional components of a contrast correcting apparatus 1 in
accordance with the second preferred embodiment, and FIG. 23 is a
flow chart that shows one portion of the operation of the contrast
correcting apparatus 1. Here, FIGS. 22 and 23 correspond to
portions of FIGS. 3 and 4 in the first preferred embodiment, and
the contrast correcting apparatus 1 in accordance with the second
preferred embodiment has the same arrangement as that of the first
preferred embodiment, except that the gray level histogram
calculation section 201 in FIG. 3 is replaced by a judgment region
calculation section 201a and that the step S13 in FIG. 4 is
replaced by a step S13a.
[0075] In the second preferred embodiment, after the pixel values
of an input image have been transformed to HSL values by the
judgment region calculation section 201a (step S12), an extraction
for the judgment region is carried out so as to make a judgment on
the scene (step S13a). The judgment region is found as a region
where pixels having gray levels within a predetermined range exist,
or as a region where pixels having hues within a predetermined
range exist. Then, based upon the judgment region thus found, the
scene judgment section 202 carries out a scene judgment, thereby
specifying scene information (step S14). Thereafter, in the same
manner as the first preferred embodiment, the amount of contrast
correction and the size of the unit regions are determined based
upon the scene information, and the contrast correction is carried
out for each of the unit regions (FIG. 4: steps S15 to S17, FIG. 5:
steps S21 to S27).
[0076] In the scene judgment (step S14), for example, in the case
when regions having gray levels higher than a predetermined gray
level are located on the periphery of an image and regions having
gray levels lower than the predetermined gray level are located in
the center of the image, it is judged that the image in question
has been picked up under a condition of back light. Moreover, in
the case when regions having gray levels lower than the
predetermined gray level are located all over the image, it is
judged that the image is a night view.
[0077] With respect to hues, they are also utilized for the scene
judgment in the same manner, and for example, in the case when a
great region is biased on red color to yellow color, it is judged
that the image shows an evening glow, and in the case when a great
region having the same color is located in the center, it is judged
that the image is a close-up shooting of a face, etc.
[0078] In this manner, the scene judgment is carried out by
extracting a judgment region in which the values of predetermined
color element, such as gray levels and hues (or saturation, RGB
values, L*a*b* values, etc.) are located within a predetermined
range. In particular, in the case when the judgment region is
great, this indicates that a region having substantially the same
gray level, hue, etc. (a so-called solid region) is great, and in
this type of images, when the size of the unit regions is
decreased, the difference in gray level transformation curves
between the unit regions appears as unevenness in gray levels in a
conspicuous manner.
[0079] Therefore, in the contrast correcting apparatus 1, when any
scene information such as back light, night view, evening glow and
close-up shooting is directed based upon the sizes of various
judgment regions, the region size determining section 204 is
allowed to set the size of the unit regions to a greater size,
independent of the amount of contrast correction.
[0080] Here, in the case of back light, the image is more
susceptible to low contrast partially; therefore, upon receipt of
scene information indicating back light, a greater value is set as
the amount of contrast correction by the contrast correction amount
determining section 203.
[0081] Moreover, in the above-mentioned operation, the size of the
unit regions is increased in accordance with the scene information
directed from the size and state of a judgment region; however, the
size of the judgment region and the size of the unit regions may be
directly correlated with each other. For example, the size of the
judgment region is utilized as an index that shows the degree of
back light, night view, evening glow and close-up shooting, and as
the judgment region is increased, the size of the unit regions may
be increased.
3. Third Preferred Embodiment
[0082] In the first and second preferred embodiments, the scene
information is automatically directed from an image state; however,
the scene information may be preliminarily provided as information
for specifying a state of an image. The following description will
discuss the third preferred embodiment in which the scene
information is preliminarily provided.
[0083] FIG. 24 is a block diagram that shows one portion of
functional components of a contrast correcting apparatus 1 in
accordance with the third preferred embodiment, and FIG. 25 is a
flow chart that shows one portion of the operation of the contrast
correcting apparatus 1. FIG. 24 shows a state in which the gray
level histogram calculation section 201 and the scene judgment
section 202, shown in FIG. 3, are omitted; and FIG. 25 shows a
state in which step S13 and step S14, shown in FIG. 4, are omitted,
and step S11a is added thereto. Other arrangements and operations
are the same as those shown in the first preferred embodiment.
[0084] As illustrated in FIG. 24, in the third preferred
embodiment, image data together with scene information are supplied
to the contrast correction amount determining section 203 and the
region size determining section 204 from the digital camera 93
(steps S11, S11a). The image data is supplied to the gray level
transformation curve forming section 205 where it is transformed to
HSL values (step S112).
[0085] In the contrast correction amount determining section 203,
the scene information (and image data, if necessary) is used to
determine the contrast correction amount (FIG. 4: step S15).
Moreover, in the region size determining section 204, based upon
the scene information given from the digital camera 93, the size of
the unit regions is determined (step S16). After the image is
divided, in the same manner as the first preferred embodiment, the
gray level transformation curves are formed for the respective unit
regions, thereby carrying out the correction of the image contrast
(step S17, FIG. 5: steps S21 to S27).
[0086] Here, scene information to be given to the contrast
correcting apparatus 1 from the digital camera 93 is mainly
classified into two types. One type includes pickup conditions of
the digital camera 93 at the time of picking up an image, such as
an exposure value (a shutter speed and a diaphragm value),
automatic ON/OFF of flash and a light metering sensor value, and
the other type includes information input into the digital camera
93 by the operator, such as photographing modes including landscape
mode, night view mode, room mode, etc., a fixed exposure value,
forced ON/OFF of flash. Although some of these are substantially
common items, these are different in that the former is scene
information determined at the time of an image pickup (or after an
image pickup), while the latter is scene information that has been
preliminarily determined before an image pickup.
[0087] Here, for example, in the case when the exposure value is
greater than the light metering sensor value, the region size
determining section 204 sets a greater size as the size of the unit
regions, or upon receipt of a night view mode, it sets a greater
size as the size of the unit regions.
[0088] In this manner, by utilizing information related to an image
pickup operation that is transmitted from the digital camera 93 as
the scene information, it is possible to realize a proper contrast
correction.
[0089] In the example shown in FIG. 24, an explanation will be
given of a case in which the scene information is transmitted from
the digital camera 93; however, the operator may set scene
information such as underexposure, overexposure, low contrast, high
contrast, etc. on the contrast correcting apparatus 1 through a key
board 11a and a mouse 11b while viewing an image displayed on the
display 91. In this case, not only for images obtained from the
digital camera 93, but also for images obtained through a scanner
or computer communications, the size of the unit regions is
determined based upon the scene information, thereby making it
possible to realize a proper contrast correction.
4. Modified Examples
[0090] The above description has discussed the preferred
embodiments, and these preferred embodiments may be modified to
various forms.
[0091] For example, in the above-mentioned preferred embodiment, an
accumulated histogram related to the gray level values of the unit
region is utilized as gray level transformation curve; however, as
is disclosed in "Fast Adaptive Contrast Enhancement Method for the
Display of Gray-Tone images" (The Institute of Electronics,
Information and Communication Engineers, Journal of reports, D-II,
vol. J77-D-II, No. 3, pp. 502-509, 1994/3) written by Naoki
Kobayashi, et al., a selection may be made from preliminarily
provided gray level transformation curves based upon the average
gray level of the unit region.
[0092] The alteration of the amount of contrast correction may be
made by simply altering the state of the gray level transformation
curve. In this case, the degree at which the high lighted portion
is raised and the shadow portion is lowered in the gray level
transformation curve corresponds to the amount of contrast
correction. Moreover, the contrast correction may be carried out by
increasing the difference between the maximum value and the minimum
value of the gray levels, and in this case, the amount of
alteration of the difference between the maximum value and the
minimum value of the gray levels corresponds to the amount of
contrast correction.
[0093] The amount of contrast correction is not necessarily set to
the same value for the respective unit regions; and for example, in
the case of back light, the amount of contrast correction may be
increased only in the center portion thereof.
[0094] The size and shape of the unit region are not intended to be
limited by the above-mentioned preferred embodiments; and any
method may be used as long as the contrast is corrected for each of
the unit regions.
[0095] In the above-mentioned first and second preferred
embodiments, upon determining the scene information from the gray
level histogram, the pixel values are transformed to HSL values;
however, they may be transformed to other values (for example,
L*a*b* values) that can specify the gray level values. In the first
preferred embodiment, the histogram by which the scene information
is directed is not limited to the gray level histogram; and a
saturation histogram, a hue histogram, or histograms for the
respective RGB values, etc. may be used. For example, scenes such
as evening glow and night view can be judged by using a hue
histogram.
[0096] In the above-mentioned preferred embodiments, the contrast
correction program 131 is installed in the computer 10 from the
recording medium 8; however, the recording medium 8 is not limited
to a portable recording medium, and a fixedly installed recording
device such as a fixed disk may be used. The recording device may
be connected to the computer 10 through a communication network
such as the Internet.
[0097] In the above-mentioned preferred embodiment, an explanation
was given of a case in which the computer 10 is mainly allowed to
function as the contrast correcting apparatus 1; however, the
entire functional components or a portion of them shown in FIG. 3
may be constituted as a dedicated electrical circuit. Moreover, the
program 131, in cooperation with another program, may allow the
computer 10 to function as the contrast correcting apparatus 1.
[0098] While the invention has been shown and described in detail,
the foregoing description is in all aspects illustrative and not
restrictive. It is therefore understood that numerous modifications
and variations can be devised without departing from the scope of
the invention.
* * * * *