JP2009212750A - Image processing apparatus and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing apparatus and method capable of generating images in which wherever the part of an input image may be, has high visibility and will not be discontinuous. <P>SOLUTION: An area dividing part 101 divides the input image into a plurality of small area images of a prescribed size, a similarity-determining part 102 determines whether the plurality of divided small area images are, respectively similar to one another, on the basis of pixel values of the small area images; a gamma conversion table generating part 104 generates a gamma conversion table, on the basis of a pixel value of an intermediate area image composed of a plurality of small area images determined as being similar to one another; an image converting part 105 generates a conversion intermediate area image. obtained by converting the intermediate area image with the gamma conversion table; and a discontinuity correcting part 106 corrects the pixel value of pixels near a boundary between the conversion intermediate area images adjacent to one another, by using a conversion table weighted by a weighting coefficient, based on the relation of the pixel to the respective conversion intermediate area images. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image processing apparatus and method for correcting an image.

  In recent years, with the widespread use of digital cameras and mobile phones with photographing functions, it has become possible for amateurs, not professionals, to easily photograph digital images. However, images taken by amateurs are more susceptible to the shooting environment and the like than those taken by professionals, and images intended by the photographer are often not taken.

  In such a case, various methods are known for converting an image to a high-quality image after shooting. For example, to improve the gradation of an image, the gradation level of the signal of the input image is converted. As a technique for performing this, a histogram equalization method is known. The histogram equalization method can efficiently increase the amount of luminance information held by an image by performing level conversion that makes the luminance histogram of the image uniform.

  Here, if such level conversion is applied to the entire screen at once regardless of the type of image, for example, the left side of the image can be converted as intended by the photographer, but the right side of the image is heavily emphasized by noise. In some cases, the output image may deteriorate depending on the part.

  As one of the solutions, an imaging apparatus is disclosed that suppresses degradation of an output image by searching for a main subject from the distribution of the histogram of the image and converting the subject so that it is most converted. (See Patent Document 1).

JP 2007-221678 A

  However, since the technique described in Patent Document 1 places importance on correction of only the main part, it may be converted to a level not desired by the user in other areas. For example, there are cases where two or more main objects are present in the image. In such a case, there is no guarantee that all the main parts are converted in the same conversion in the direction of increasing visibility, and in many cases the image quality of an image including one main subject is improved. However, there is a problem that the image quality of an image including another subject deteriorates.

  In addition, in order to improve visibility, instead of performing uniform image conversion on the entire image, if image conversion is performed according to the image quality of the area of the image, areas that have undergone different image conversion due to differences in image conversion processing There is a problem that the boundary of the image becomes a discontinuous image.

  The present invention has been made in view of the above, and an object of the present invention is to provide an image processing apparatus and method that can generate an image that has good visibility and is not discontinuous in any part of an input image. To do.

  In order to solve the above-described problems and achieve the object, the present invention provides an area dividing unit that divides an input image into a plurality of small area images of a predetermined size, and a division based on the pixel values of the small area images. A similarity determination unit that determines whether or not each of the plurality of small region images is similar to each other, and a gamma conversion table based on the pixel values of the middle region image that is composed of the plurality of small region images determined to be similar to each other. A pixel value of a pixel in the vicinity of the boundary between the converted middle region images adjacent to each other, a gamma conversion table generating unit to generate, an image converting unit that generates the middle region image converted from the middle region image by the gamma conversion table And a correction unit that corrects the image using a conversion table weighted by a weighting coefficient based on the relationship between the pixel and each of the conversion-in-progress region images.

  Further, according to the present invention, an area dividing unit that divides an input image into a plurality of small area images of a predetermined size and a plurality of divided small area images are similar to each other based on pixel values of the small area images. A similarity determination unit that determines whether or not, a gamma conversion table generation unit that generates a gamma conversion table based on pixel values of a middle region image composed of a plurality of small region images determined to be similar to each other, and A correction gamma conversion table generating unit that generates a correction gamma conversion table using a gamma conversion table weighted by a weighting coefficient based on a relationship between the middle region image corresponding to the gamma conversion table and each of the middle region images; An image conversion unit that converts the region image using the generated correction gamma conversion table.

  Further, the present invention is an image processing method executed by the image processing apparatus, wherein the region dividing unit divides the input image into a plurality of small region images of a predetermined size, and the similarity determining unit includes: A similarity determination step for determining whether or not each of the plurality of divided small area images is similar to each other based on a pixel value of the small area image, and a plurality of gamma conversion table generation units determined to be similar to each other A gamma conversion table generating step for generating a gamma conversion table based on the pixel value of the middle area image made up of the small area image, and a converted middle area image obtained by converting the middle area image by the gamma conversion table by the image conversion unit. And an image conversion step for generating a pixel value of a pixel in the vicinity of a boundary between the conversion target region images adjacent to each other based on a relationship between the pixel and the conversion target region image. Characterized in that it comprises a correction step of correcting by using a conversion table weighted by a factor.

  The present invention is also an image processing method executed by the image processing apparatus, wherein the region dividing unit divides the input image into a plurality of small region images of a predetermined size, and the similarity determining unit includes: A similarity determination step for determining whether or not each of the plurality of divided small area images is similar to each other based on the pixel value of the small area image, and a plurality of gamma conversion table generation units determined to be similar to each other A gamma conversion table generating step for generating a gamma conversion table based on the pixel value of the middle area image made up of the small area image, and a correction gamma conversion table generating unit includes a middle area image corresponding to the generated gamma conversion table, A correction gamma conversion table that generates a correction gamma conversion table using a gamma conversion table weighted by a weighting factor based on the relationship with each of the middle region images. And Bull generating step, the image conversion unit, said in region image, and an image conversion step of converting the generated said correction gamma conversion table, characterized in that it comprises a.

  According to the present invention, optimal image conversion is performed for each similar region, and pixels in the vicinity of the boundary of the region are corrected. Therefore, an image that is highly visible in any part of the input image and does not become discontinuous is generated. There is an effect that can be.

  In addition, according to the present invention, since image conversion that is appropriate for each similar region and does not cause discontinuity in the image near the boundary of the region is performed, visibility is good in any part of the input image, and discontinuity is achieved. There is an effect that an image that does not become can be generated.

  Exemplary embodiments of an image processing apparatus and method according to the present invention will be explained below in detail with reference to the accompanying drawings. The present invention is not limited to these embodiments.

(First embodiment)
A first embodiment will be described with reference to the accompanying drawings. First, a configuration example of an image processing apparatus to which the present invention is applied will be described.

  FIG. 1 is a block diagram illustrating a configuration of the image processing apparatus according to the first embodiment. The image processing apparatus 100 according to the present embodiment includes an area dividing unit 101, a similarity determining unit 102, a middle region generating unit 103, a gamma conversion table generating unit 104, an image converting unit 105, and a discontinuous correcting unit 106. And mainly.

  The area dividing unit 101 divides the input image into small area images. There is no particular limitation on the image dividing method. However, the similarity determination unit 102, which will be described later, uses a small region image that is large enough to calculate the similarity.

  FIG. 2 is an explanatory diagram illustrating an example in which an input image is divided into small region images. Here, for explanation, the origin 22 is provided at the upper left corner of the image to be processed, and the x axis is set in the horizontal direction and the y axis is set in the vertical direction. In the example illustrated in FIG. 2, the input image 21 is divided into eight equal parts in the x-axis direction and the y-axis direction, and divided into 64 small region images 23. Note that FIG. 2 is an example, and when the input image is actually divided into small region images, it is not necessary to equally divide the region, and the region may not be rectangular. Further, there is no restriction that the input image is all uniformly divided into the same shape. For example, a region of particular interest in the image may be divided finely and roughly divided in other ranges.

  The similarity determination unit 102 calculates the similarity of the pixel values between the small region images divided by the region dividing unit 101, and determines whether the small region images are similar from the calculated similarity. Here, the degree of similarity is an index that indicates how similar pixel values (for example, brightness components of the image such as luminance and RGB values) are compared between the small region images. The means for obtaining the similarity may be any method as long as it is a generally known method of comparing distributions such as luminance.

  For example, the following methods (1) to (9) are available as means for obtaining the similarity of images to be compared using luminance. (1) The degree of similarity is increased when the difference between the average values of the contrasted images is small. (2) The degree of similarity is increased when the difference between the maximum values of the brightness of the images to be compared is small. (3) When the difference in the distance between the minimum values of the brightness of the images to be compared is small, the similarity is increased. (4) When the difference in distance between the maximum value and the minimum value of the brightness of the contrasted images is small, the similarity is increased. (5) When the difference in luminance dispersion is small, the similarity is increased. (6) The degree of similarity is increased when the difference in the level of the mode of brightness of the compared images is small. (7) The degree of similarity is increased when the difference in the median level of the brightness of the images to be compared is small. (8) The brightness histograms of the images to be compared are overlapped, and the similarity is increased when the overlapping area is large. (9) The shift histograms obtained by shifting the mode of the histograms of the brightness of the contrasted images to 0 are overlapped, and the similarity is increased when the overlap area is large.

  Further, not only the comparison using the luminance, but the same method can be considered as long as the comparison is performed using a value such as an RGB value that shows the brightness. Moreover, you may combine several calculation methods of the similarity mentioned above. In addition to this, any index may be used as long as it indicates the luminance distribution of the image, such as color, edge distribution, and corner point. The similarity obtained in this way is compared with a preset threshold value to determine whether or not the small areas are similar. Note that the small areas to be compared may or may not be adjacent to each other.

  FIG. 3 is a flowchart illustrating an example of a similarity determination processing procedure performed by the similarity determination unit 102. This is a combination of the above methods (6) and (9) for obtaining the similarity.

  First, the similarity determination unit 102 generates a luminance histogram from an image of a small area whose similarity is to be obtained (step S301). 4A and 4B are explanatory diagrams illustrating an example of a histogram of luminance values generated from a small region image. In the histogram shown in FIG. 4A, the peak 41 is shown by the luminance value K1, and in the histogram shown in FIG. 4B, the peak 42 is shown by the luminance value K2.

  The similarity determination unit 102 calculates the mode value of the generated histogram (step S302). In the image shown in FIG. 4A, the luminance value K1 is calculated as the mode value of the histogram, and in the image shown in FIG. 4B, the luminance value K2 is calculated as the mode value of the histogram. Next, the similarity determination unit 102 determines whether or not the mode value difference is smaller than the threshold (step S303). If it is determined that the difference between the mode values is smaller than the threshold value (step S303: Yes), the similarity determination unit 102 superimposes histograms obtained by shifting the mode value of the histogram of luminance values to 0, and calculates the area of overlap. (Step S304). FIG. 5 is an explanatory diagram showing an example of a histogram in which the mode value of the histogram of luminance values is shifted to 0 and superimposed. The area of the portion 51 where the two histograms overlap as shown in FIG. 5 is obtained by integrating the smaller vertical axis values of the two histograms with the same horizontal axis values.

  The similarity determination unit 102 determines whether or not the overlapping area is larger than a predetermined value (step S305). When it is determined that the overlapping area is larger than the predetermined value (step S305: Yes), the similarity determination unit 102 determines that the compared small area images are similar (step S306).

  If it is determined in step S303 that the mode difference is greater than or equal to the threshold (step S303: No), or if it is determined in step S305 that the overlap area is less than or equal to a predetermined value (step S305: No), the similarity determination unit 102 determines that the compared small area images are not similar (step S307).

  Returning to FIG. 1, the middle region generation unit 103 will be described. Based on the determination result of whether or not two or a plurality of small region images are similar by the similarity determination unit 102, the middle region generation unit 103 creates a new region in which the small regions of similar images are combined into one. Create a region. Hereinafter, an image of a new area obtained by combining two or a plurality of small area images into one is referred to as a middle area image. As long as the middle region image is a small region image determined to be similar by the similarity determination unit 102, the middle region image may be adjacent to or separated from the target small region image.

  FIG. 6 is an explanatory diagram illustrating an example of a small region image determined to be similar by the similarity determination unit 102. In the example illustrated in FIG. 6, as a result of determining whether or not each of the small region images 61 is similar, a bidirectional arrow 62 indicating similarity is represented between the similar small region images. In the case shown in FIG. 6, the similarity determination unit 102 determines similarity for four region images (upper, lower, left, and right) for a certain small region image.

  FIG. 7 is an explanatory diagram illustrating an example of a result of the middle region generation unit 103 generating a middle region image from the similarity determination result. As shown in FIG. 7, a plurality of similar small area images shown in FIG. 6 are grouped as a middle area image. In the example illustrated in FIG. 7, a middle region image 71 is generated for each subject such as a road surface, sky, mountain, building, person, and car.

  The gamma conversion table generation unit 104 generates a signal conversion gamma conversion table for each of the middle region images generated by the middle region generation unit 103 based on the pixel values of the middle region image. As a method for generating a gamma conversion table for signal conversion, any generally known image conversion method may be used. For example, conversion methods such as a histogram equalization method and output = (input) ^ γ can be mentioned.

  The image conversion unit 105 performs signal conversion on each middle region image using the signal conversion gamma conversion table generated by the gamma conversion table generation unit 104. In this embodiment, a case where signal conversion is performed using a signal conversion gamma conversion table generated by the gamma conversion table generation unit 104 will be described. However, the present invention is not limited to signal conversion using a gamma conversion table. A completely different image conversion method may be used for each middle region image.

  The discontinuity correction unit 106 corrects pixels near the boundary of the middle region image converted by the image conversion unit 105. In the image converted by the image conversion unit 105, the image conversion differs for each middle region image, so that an image having a discontinuous gradation may occur at the boundary of the middle region image. Therefore, the discontinuity correction unit 106 performs processing for removing this discontinuous boundary. In the boundary removal, a discontinuous image is corrected by using a generally known deblocking method for the boundary portion or the entire image. As an example, image discontinuity can be removed by blurring the boundary using a Gaussian filter or the like.

As another method, a discontinuous image can be corrected by the following method. A case where a certain pixel A (coordinate (X, Y), pixel value P) is converted will be described as an example. An input / output conversion table in the middle region image i (i = 1, 2,..., I, I: the number of generated middle region images) generated by the gamma conversion table generation unit 104 is represented by f _i (P ). At this time, the pixel value P ′ after conversion in the pixel A is obtained by weighted linear interpolation represented by the following formula (1).

Here, W _i (P, X, Y) is the pixel value of the pixel A to be converted, the position of the pixel A in the input image, the distance between the pixel A and the middle region image, the area of each middle region image, the middle region This is a weighting coefficient determined by the luminance distribution of each image.

  This weight coefficient is determined by the relationship between the pixel to be converted and each of the middle region images. Various design methods can be considered for the weighting factor. As a basic design policy, it is preferable to use a weighting factor by using the distance between the middle region image and the pixel A and the area of the middle region image.

  For example, (1) weighting is performed according to the distance between the center of gravity G of the middle region image and the pixel A. Specifically, the weighting factor is increased as the distance between the center of gravity G of the middle region image and the pixel A is shorter, and the weighting factor is decreased as the distance is longer. (2) Weighting is performed according to the area of the middle region image. Specifically, according to the size of each area of the middle region image, the weighting factor is increased as the area is larger, and the weighting factor is decreased as the area is smaller. In addition, weighting is performed with a weighting coefficient using the area of the middle region image included in the circle within the radius R with the pixel A as the center. In the case of the example (1), the minimum distance between the pixel A and the boundary of each middle region image may be used. In the case of the example (2), a square or a rectangle may be used instead of a circle. Furthermore, instead of or in addition to the methods (1) and (2), a weighting coefficient corresponding to the luminance distribution of the middle region image may be added. For example, the weight coefficient is increased as the number of pixels having a luminance value in a predetermined range is increased in the middle region image, and the weight coefficient is decreased as the number of pixels having a luminance value in the predetermined range is decreased. By these methods, the discontinuity at the boundary can be corrected.

  In this way, by performing gamma conversion for each middle region image obtained by dividing the input image and collecting similar images, optimal gamma conversion can be performed for each part of the input image, and further, the middle region image In order to perform correction to reduce the discontinuity of the image for the pixels near the boundary of the converted middle-region image by correcting the pixels near the boundary of the image in the conversion table, any part of the input image There is an effect that it is possible to generate an image that has good visibility and is not discontinuous.

(Second Embodiment)
A second embodiment will be described with reference to the accompanying drawings. A configuration example of an image processing apparatus to which the present invention is applied will be described. FIG. 8 is a block diagram illustrating a configuration of an image processing apparatus according to the second embodiment.

  The image processing apparatus 200 according to the present embodiment includes an area dividing unit 101, a similarity determining unit 102, a middle region generating unit 103, a gamma conversion table generating unit 104, a corrected gamma conversion table generating unit 207, and a corrected image. A conversion unit 208 is mainly provided. Here, the configurations and functions of the region dividing unit 101, the similarity determining unit 102, the middle region generating unit 103, and the gamma conversion table generating unit 104 are the same as those in the first embodiment, and thus the description thereof is omitted. .

  The correction gamma conversion table generation unit 207 corrects each of the signal conversion gamma conversion tables generated by the gamma conversion table generation unit 104 so that no discontinuity occurs in the boundary portion between the middle region images. Generate a conversion table. As described above, the corrected gamma conversion table generation unit 207 has a discontinuity of gradation between middle region images that occurs when image conversion is performed using the gamma conversion table generated by the gamma conversion table generation unit 104 as it is. Is suppressed by correcting the gamma conversion table. Here, the corrected gamma conversion table is referred to as a corrected gamma conversion table.

  Specifically, the gamma conversion table of a certain middle area image is smoothed using the output values at the same level of the gamma conversion table of the middle area image and other middle area images of interest at each level, Find the corrected output value. A correction gamma conversion table is generated by obtaining at all levels. By performing this operation for all of the middle region images, gradation discontinuity between the middle region images is suppressed.

There are various methods for smoothing the gamma conversion table of a certain middle region of interest. A case where a corrected gamma conversion table F _B (X) for a certain middle region image B is generated will be described as an example. If the gamma conversion table of each middle region image is fi (X) (i = 1, 2,..., I, I: number of generated middle region images), F _B (X) is expressed by the following formula ( It is calculated | required by the weighted linear interpolation represented by 2).

  Here, W (B, i) is a weighting factor determined by the position, area, distance, luminance distribution, etc. of the middle region image B and the middle region image i. As a basic design policy, it is preferable to use a distance between middle region images as a weighting factor.

As a design method of the weighting factor, for example, there is a method of designing according to the distance between the center of gravity G _B of the middle region image and the center of gravity G _i of the middle region image i. The weighting factor is increased as the distance between the center of gravity G _B of the middle region image and the center of gravity G _i of the middle region image i is shorter, and the weighting factor is decreased as the distance is longer. Furthermore, instead of or in addition to these distances, the weighting factor may be corrected by the size of the area of the middle region image, the similarity of the luminance distribution of the middle region image, or the like. For example, the weighting factor is increased as the similarity of the luminance distribution between the middle region image B and the middle region image i is increased, and the weighting factor is decreased as the similarity is decreased. As a method for calculating the similarity, the method exemplified in the similarity determination unit 102 may be used. By correcting the gamma conversion table by these methods and using the corrected gamma conversion table corrected, the discontinuity at the boundary can be corrected.

  The corrected image conversion unit 208 performs signal conversion on each of the middle region images using the correction gamma conversion table for signal conversion generated by the correction gamma conversion table generation unit 207.

  In this way, a gamma conversion table is generated for each intermediate region image in which similar images are divided by dividing the input image, and the relationship between the intermediate region image corresponding to the generated gamma conversion table and other intermediate region images By generating a correction gamma conversion table using a gamma conversion table weighted with a weighting coefficient based on the above, and converting the middle region image by the correction gamma conversion table, a better gamma conversion is performed for each part of the input image, In addition, since each of the middle region images is subjected to gamma conversion while reducing discontinuity in image quality with the adjacent middle region image, it is possible to generate an image that is highly visible in any part of the input image and does not become discontinuous. There is an effect.

  FIG. 9 is an explanatory diagram illustrating an example of a hardware configuration of the image processing apparatus according to the above-described embodiment. As shown in FIG. 9, the image processing apparatuses (100 and 200) according to the above-described embodiments have a hardware configuration according to a ROM 902 that stores an image processing program for performing the above-described processing, and a program in the ROM 902. A CPU 901 that controls each unit of the image processing apparatus (100 and 200), a RAM 903 that is a data storage area, a communication I / F 904 that communicates by connecting to a network to input and output moving images, and each unit And a bus 905 to be connected.

  The image processing program may be provided by being recorded in a computer-readable recording medium such as an optical disk, a magnetic disk, and an SSD (Solid State Disk) in an installable or executable format file.

  In this case, the image processing program is loaded on the RAM 903 by being read from the recording medium and executed by the image processing apparatuses (100 and 200), and each unit described in the software configuration is generated on the RAM 903. It has become.

  Further, the image processing program according to the above-described embodiment may be provided by being stored on a computer connected to a network such as the Internet and downloaded via the network.

  Note that the present invention is not limited to the above-described embodiments as they are, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. Various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the above-described embodiments. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

1 is a block diagram illustrating a configuration of an image processing apparatus according to a first embodiment. It is explanatory drawing which shows an example which divided | segmented the input image into the small area | region. It is a flowchart which shows an example of the similarity determination process procedure which a similarity determination part performs. It is explanatory drawing which shows an example of the histogram of the luminance value produced | generated from the small area image. It is explanatory drawing which shows an example of the histogram of the luminance value produced | generated from the small area image. It is explanatory drawing which shows an example of the histogram which shifted the mode value of the histogram of a luminance value to 0, and was superimposed. It is explanatory drawing which shows an example of the small area | region determined with the similarity determination part being similar. It is explanatory drawing which shows an example of the result as which the middle region production | generation part produced | generated the middle region from the result of similarity determination. It is a block diagram which shows the structure of the image processing apparatus concerning 2nd Embodiment. It is explanatory drawing which shows an example of the hardware constitutions of the image processing apparatus concerning embodiment mentioned above.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 200 Image processing apparatus 101 Area division part 102 Similarity determination part 103 Middle area | region production | generation part 104 Gamma conversion table production | generation part 105 Image conversion part 106 Discontinuous correction part 207 Correction | amendment gamma conversion table generation part 208 Correction | amendment image conversion part

Claims (9)

An area dividing unit that divides an input image into a plurality of small area images of a predetermined size;
A similarity determination unit that determines whether or not each of the plurality of divided small region images is similar to each other based on the pixel value of the small region image;
A gamma conversion table generating unit that generates a gamma conversion table based on pixel values of a middle region image composed of a plurality of small region images determined to be similar to each other;
An image conversion unit that generates a converted middle region image obtained by converting the middle region image with the gamma conversion table;
A correction unit that corrects the pixel values of the pixels in the vicinity of the boundary between the conversion-target region images adjacent to each other using a conversion table weighted by a weighting factor based on the relationship between the pixel and each of the conversion-target region images;
An image processing apparatus comprising:   The correction unit corrects a pixel value of a pixel near a boundary of the mid-conversion image using a conversion table weighted by increasing a weighting factor as the distance between the pixel and the mid-conversion image is shorter. The image processing apparatus according to claim 1, wherein:   The correction unit corrects a pixel in the vicinity of a boundary of the mid-conversion image using a conversion table weighted with a weighting factor corresponding to the size of each area of the mid-conversion image. The image processing apparatus according to claim 1 or 2.   The correction unit corrects a pixel in the vicinity of a boundary of the mid-conversion image using a conversion table weighted with a weighting coefficient corresponding to a luminance distribution of each mid-conversion image. The image processing apparatus according to claim 2. An area dividing unit that divides an input image into a plurality of small area images of a predetermined size;
A similarity determination unit that determines whether or not each of the plurality of divided small region images is similar to each other based on the pixel value of the small region image;
A gamma conversion table generating unit that generates a gamma conversion table based on pixel values of a middle region image composed of a plurality of small region images determined to be similar to each other;
A corrected gamma conversion table generating unit that generates a corrected gamma conversion table using a middle region image corresponding to the generated gamma conversion table and a gamma conversion table weighted by a weighting factor based on a relationship between each of the middle region images; ,
An image conversion unit for converting the middle region image with the generated correction gamma conversion table;
An image processing apparatus comprising:   The correction gamma conversion table generation unit is configured to perform gamma conversion in which a weighting factor is increased and weighted as the distance between the intermediate region image corresponding to the generated gamma conversion table and the intermediate region image other than the intermediate region image is shorter. The image processing apparatus according to claim 5, wherein a correction gamma conversion table is generated using the table.   The corrected gamma conversion table generation unit weights a gamma weighted with a weighting factor according to the similarity of luminance distribution between the middle area image corresponding to the generated gamma conversion table and a middle area image other than the middle area image. The image processing apparatus according to claim 5, wherein a correction gamma conversion table is generated using the conversion table. An image processing method executed by an image processing apparatus,
An area dividing unit that divides the input image into a plurality of small area images of a predetermined size; and
A similarity determination step for determining whether or not each of the plurality of divided small region images is similar to each other based on a pixel value of the small region image;
A gamma conversion table generating unit that generates a gamma conversion table based on pixel values of a middle region image composed of a plurality of small region images determined to be similar to each other;
An image conversion step for generating a converted middle region image obtained by converting the middle region image with the gamma conversion table;
A correction step in which the correction unit corrects the pixel values of the pixels in the vicinity of the boundary between the conversion-in-progress region images adjacent to each other using a conversion table weighted by a weighting coefficient based on the relationship between the pixel and each of the conversion-in-progress region images. When,
An image processing method comprising: An image processing method executed by an image processing apparatus,
An area dividing step of dividing the input image into a plurality of small area images of a predetermined size;
A similarity determination step for determining whether or not each of the plurality of divided small region images is similar to each other based on a pixel value of the small region image;
A gamma conversion table generating unit that generates a gamma conversion table based on pixel values of a middle region image composed of a plurality of small region images determined to be similar to each other;
A corrected gamma conversion table generation unit generates a corrected gamma conversion table using a gamma conversion table weighted by a weighting coefficient based on a relationship between the generated middle area image corresponding to the generated gamma conversion table and each of the middle area images. A correction gamma conversion table generation step to perform,
An image conversion step in which the image conversion unit converts the middle region image with the generated correction gamma conversion table;
An image processing method comprising:

Images