JP2006146646A - Method and apparatus for processing photographic image - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a photographic image processor capable of suitably executing correction processing for the whole area of an input image while utilizing the effects of a lookup table which can be used for high-speed processing and capable of obtaining an output image of high quality. <P>SOLUTION: The photographic image processor for generating an output image by correcting an inputted original image is provided with: an area division means 50a for dividing the original image into a first area including a main subject and a second area other than the first area; an LUT selection means 50d for selecting an optimum LUT based on an image characteristic of each divided area; and an LUT processing means 50c for executing correction processing by using LUTs respectively selected for the first area and the second area. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a photographic image processing method and photographic image processing for correcting gradation and the like on an original image read by a film scanner from a developed silver halide photographic film or an original image taken by a digital still camera. Relates to the device.

  In the case where image information input as a digital signal is output to a photosensitive material, photographic paper, or the like, in order to obtain a high-quality output image, it is necessary to perform at least correction processing on the input image. For example, contrast adjustment and sharpness enhancement. For this correction processing, LUT (look-up table) processing, filter processing, or the like is used.

  Since the correction processing by the filter processing calculates the correction value using a plurality of data around the correction target data, it is very effective for advanced correction processing technology, mainly for enhancing sharpness and the like. Although it is preferable, since it is necessary to perform complex matrix calculation processing for a plurality of pixel data, the calculation speed takes time, and improvement of the correction processing speed is a problem.

  On the other hand, the correction process by the LUT process is effective only for a simple correction process because the conversion process is performed simply based on the LUT in which the output value for the input value of the correction target data is individually determined. Although it is mainly suitable for adjustment of density and contrast, it is not suitable for advanced correction processing technology because it is not related to peripheral data of correction target data.

  For example, when the input image is a photographic image taken with a strobe in a dark environment, the background becomes black when correction processing is performed using LUT data in which the density and contrast of the subject are appropriate for density and contrast adjustment. On the contrary, if the correction process is performed using LUT data in which the background density and contrast are appropriate, the subject may fly white. Also, if LUT data with appropriate subject and background densities is set, the overall contrast may be weak, resulting in a low-quality output image. Further, there is a similar problem even in a backlight scene. In a projection exposure type photographic processing apparatus, dodging is performed to appropriately adjust the density and contrast of both the subject and the background.

Further, in the medical image diagnostic apparatus, the image memory for storing the image, the means for converting the image in the image memory into an arbitrary gradation using a lookup table, and the gradation conversion means convert the gradation. An image processing apparatus comprising: a display control device for displaying a displayed image; and an operation console for inputting an image displayed on the display control device and a specific area of an image memory corresponding to the image. A look-up table used for image tone conversion on the screen of the display control device for the specific area, a look-up table used for tone conversion of the entire image, and a means for synthesizing these look-up tables. A technique for performing gradation conversion using the look-up table has been proposed.
Japanese Patent Laid-Open No. 06-259543

  In the above-described prior art, in the composition of the look-up table, appropriate gradation conversion is possible for the specific area designated for input, but sufficient gradation is not ensured in other areas, There is a problem that it is difficult to apply to a general photographic image taken under various conditions.

  In view of the above-described conventional problems, the present invention can appropriately perform correction processing in the entire area of an input image while taking advantage of a good look-up table capable of high-speed processing, and can obtain a high-quality output image. It is an object of the present invention to provide a photographic image processing method and apparatus.

  In order to achieve the above-mentioned object, the first characteristic configuration of the photographic image processing method according to the present invention is that the input original image is corrected and output as described in claim 1 of the claims. A photographic image processing method for generating an image, wherein the original image is divided into a first region including a main subject and a second region other than the first region, and the first region and the second region are different. And a LUT processing step for performing correction processing using the LUT.

  According to the above-described configuration, correction processing can be performed at high speed using different LUTs for the first region including the main subject detected from the original image in the region dividing step and the second region other than the first region. By performing the correction process using an appropriate LUT, both the first area and the second area are appropriately corrected, and as a result, a high-quality output image can be obtained.

  In the second feature configuration, as described in claim 2, in addition to the first feature configuration described above, the LUT processing step performs a correction process based on an LUT corresponding to an image characteristic of each divided region. Therefore, by determining the LUT applied to each area divided in the area dividing step according to the image characteristics of the area, appropriate image correction can always be performed.

  In the third feature configuration, as described in claim 3, in addition to the first or second feature configuration described above, the LUT is selected from a plurality of preset LUTs. If an LUT corresponding to a plurality of original images having different characteristics is prepared in advance, an appropriate LUT can be quickly selected according to the image characteristics of the divided area and can be immediately processed.

  In the fourth feature configuration, in addition to the first or second feature configuration described above, the LUT is generated based on the image characteristics of each divided region. With this configuration, the LUT for correcting an image can be generated to the most appropriate value for each divided region.

  In the fifth feature configuration, as described in claim 5, in addition to any of the first to fourth feature configurations described above, the region dividing step is based on a correlation of pixel data of an original image. When dividing the first area including the main subject and the other second area, a set of pixel groups having a strong correlation and a set of pixel groups having a weak correlation are obtained. By dividing, the area can be appropriately divided.

  In the sixth feature configuration, as described in claim 6, in addition to any one of the first to fourth feature configurations described above, the region dividing step performs pattern recognition using a neural network from an original image. The region detected by the step is divided as the first region, and when the first region including the main subject and the other second region are divided into regions, the main subject is included by pattern recognition using a neural network. The first region can be accurately extracted.

  In the seventh feature configuration, as described in claim 7, in addition to any of the first to fourth feature configurations described above, the region dividing step detects a human face region from the original image. The area including the detected face area is divided as the first area, and when the main subject is a person, the area can be easily divided by detecting the area including the face area. . Here, in order to detect the face area, a color detection method for recognizing skin color or lip color, a pattern recognition method for detecting the contour shape of the skin color area, and various other known techniques can be used. .

  In the eighth feature configuration, in addition to any one of the first to seventh feature configurations described above, the first region and the second region are provided after the LUT processing step. The smoothing process step for performing the smoothing process on the boundary area of each of them is provided, and unnatural edges generated at the boundaries of the areas corrected by different LUTs can be removed.

  The first characteristic configuration of the photographic image processing apparatus according to the present invention is the photographic image processing apparatus for generating an output image by correcting the input original image, as described in claim 9, Area dividing means for dividing the image into a first area including a main subject and a second area other than that, and LUT processing means for performing correction processing using different LUTs for the first area and the second area, respectively. It is in the point.

  In the second feature configuration, as described in claim 10, in addition to the first feature configuration described above, the LUT processing unit performs correction processing based on an LUT corresponding to an image characteristic of each divided region. It is in.

  In the third feature configuration, as described in claim 11, in addition to the first or second feature configuration described above, the region dividing unit is detected from the original image by pattern recognition using a neural network. In other words, the area is divided as the first area.

  In the fourth feature configuration, as described in claim 12, in addition to the first or second feature configuration described above, the region dividing unit detects a human face region from an original image, and is detected. An area including the face area is divided as the first area.

  The fifth feature configuration includes the first region of the image processed by the LUT processing means in addition to any one of the first to fourth feature configurations described above. And smoothing processing means for performing a smoothing process on the boundary region of the second region.

  As described above, according to the present invention, a photographic image capable of appropriately performing a correction process in the entire area of the input image and obtaining a high-quality output image while taking advantage of a lookup table capable of high-speed processing. A processing method and apparatus thereof can be provided.

  The photographic image processing apparatus according to the present invention will be described below. As shown in FIGS. 1 and 2, the digital photographic image processing apparatus includes a film scanner 1a that reads a photographic image from a filmed negative film (hereinafter simply referred to as “film”) F, and a digital still camera. The image data input unit 1 includes a media driver 1b corresponding to various card-type memories M such as an SD card or a memory stick in which captured image data is stored, and a hard disk or the like that stores the input image data. An image data storage unit 2, a monitor 3 for displaying an image based on the image data, an operation input unit 4 having a keyboard and a mouse, and the operation input unit 4 for the image displayed on the monitor 3 The image processing unit 5 that edits the image data based on various operations, and the photographic paper P is exposed based on the data after the image processing. A photographic printer 6 for generating a photographic print Te, the system controller 7 controls the individual blocks of the above as a system (hereinafter referred to as "the controller 7".) Configured with a.

  The film scanner 1 a is recorded on the illumination optical system 10 that irradiates the illumination light for reading onto the film F, the film transport unit 11 that transports the film F, and the film F transported by the film transport unit 11. An image reading unit 12 for reading the frame image, and a scanner control unit 13 for controlling an image reading process by the illumination optical system 10, the film transport unit 11, and the image reading unit 12.

  The illumination optical system 10 includes a rod-shaped halogen lamp 10a arranged along a main scanning direction orthogonal to a sub-scanning direction (indicated by an arrow in FIG. 2) that is a conveyance direction of the film F, and A light control filter 10b that adjusts the color distribution of the light beam, a cylindrical lens 10c that collects the light beam in a slit shape, a diffusion plate 10d that makes the intensity distribution uniform, and a narrow slit 10e are configured. The

  The film transport unit 11 is driven by a transport motor (not shown), and includes a plurality of transport roller pairs 11a that transport the long film F toward the film projection unit directly below the slit 10e at a predetermined speed. The

  The image reading unit 12 includes a condenser lens, a CCD line sensor, a sample hold circuit, an A / D converter, and the like, and slit light from the illumination optical system 10 that has passed through the film F is collected. An analog signal formed on the CCD line sensor by the optical lens and read by the CCD line sensor is converted to digital data by the A / D converter. The CCD line sensor is composed of three lines of line sensors each provided with a color filter that selectively transmits light of each of the R component, G component, and B component of the transmitted light of the film. As the film F is conveyed, each frame image is separated into R component, G component, and B component color components and read.

  The photographic printer 6 includes a paper magazine 60 in which roll-shaped photographic paper P is accommodated, a plurality of photographic paper transport rollers 61 that pull out and transport the photographic paper P from the paper magazine 60, and a motor that drives the transport roller 61. 62, a fluorescent beam type print head 63 that exposes the photosensitive surface of the conveyed photographic paper P, a development processing unit 64 that performs development, bleaching, and fixing processes on the exposed photographic paper P, and development processing. The photographic paper P drawn out from the paper magazine 60 includes a drying unit 65 that transports the photographic paper P while drying, and a discharge unit 66 that discharges the dried photographic paper P as a final print. The paper is cut into a predetermined print size by a cutter (not shown) disposed either before or after the development processing and output to the discharge unit 66.

  The print head 63 includes three rows of a red light emission block, a green light emission block, and a blue light emission block in which phosphor elements each having a lens and a color filter mounted on a phosphor whose light emission is controlled by adjusting a grid voltage are arranged in the main scanning direction. The photographic image is exposed on the photographic paper P by controlling the driving based on the R component, G component, and B component pixel data of the image data read and edited through the film scanner 1a and the like. Is done.

  The controller 7 includes a CPU, a ROM, a RAM used as a data processing area, a RAM used for image data editing, hardware including peripheral circuits, and a CPU stored in the ROM. If it is divided into functional blocks related to the present invention and is composed of software consisting of programs to be executed, as shown in FIG. 3, a graphic including soft switches and the like on a window displayed on the monitor 3 A graphic user interface unit (hereinafter referred to as “GUI unit”) 7 a that generates a control command corresponding to a user operation by the operation input unit 4 via the operation screen display or the graphic operation screen, and the monitor 3 Are read from the video memory 7b and the image data storage unit 2. An image processing memory 7d to which the output image is loaded and subjected to various types of image processing; an image file editing unit 7c for editing the processed image data into an image file for writing to a medium such as a CD-R; A print data conversion unit 7e for converting the image processed data into output data corresponding to the photo printer 6 is provided.

  The image processing unit 5 includes software that edits the target image with a predetermined algorithm and hardware including an image processor. The image processing unit 5 performs density and gradation on the image data loaded in the image processing memory 7d. Gradation correction processing means 50 that performs correction, sharpening processing means 51 that performs sharpening processing, color correction means 52 that adjusts color balance for density, gradation correction, or sharpened image data It is configured with.

  The gradation correction processing means 50 detects the first area including the main subject (for example, a person) from the original image of each frame developed in the image processing memory 7d, and the first area and the other second area ( For example, an area dividing unit 50a for dividing a background image area), an area information generating unit 50b for generating area information indicating image characteristics of each of the divided first area and second area, and for gradation correction LUT data storage means 50f storing a plurality of LUT data, LUT selection means 50e for selecting an LUT suitable for the respective area information from the LUT data storage means 50f, and generating LUT data as necessary LUT data generation means 50f for performing the LUT processing for the first area and the second area, and a LUT processing means 50c for performing the LUT processing after the LUT processing. It is constituted by a smoothing processing unit 50g for smoothing the boundary area between the region and the second region.

  When a film reading command is sent from the controller 7, the scanner control unit 13 turns on the halogen lamp 10a and then drives the film transport unit 11 to transport the film F at a predetermined speed in the sub-scanning direction. Then, the frame images recorded on the film are sequentially read by the image reading unit 12, the read image data is transmitted to the controller 7, and the image data transmitted from the scanner control unit is stored in the image data storage unit 2. Is done.

  When a medium is inserted into the media driver 1 b under the control of the controller 7, image data stored in the medium is read and stored in the image data storage unit 2.

  The image data stored in the image data storage unit 2 is subjected to predetermined image processing by the image processing unit 5, and the print data converted by the print data conversion unit 7 e is output to the photo printer 6. In the photo printer 6, the print head 63 is driven based on the input print data, and the photographic paper P exposed by the print head 63 is developed and output as a photo print.

  More specifically, in the photographic image processing apparatus, when a photographic print is generated based on an image input from the film scanner 1a, a film image is first read at a low resolution by a pre-scan operation and displayed on the monitor 3. The image quality of each frame image is verified together with the setting processing of the print size and the number of prints by the operator via the operation input unit 4 for the displayed frame image, and gradation correction, color correction, etc. are performed as necessary. Image correction processing is performed.

  This verification operation is carried out in an interactive manner between the controller 7 and the operator, and in addition to the setting of the number of prints described above, a series of operations in which the image processing conditions are determined by the image processing unit 5, sharpening processing, color correction A series of image processing such as processing and gradation correction processing is automatically executed under predetermined conditions, and whether or not the correction result is appropriate is evaluated by an operator, and gradation correction is insufficient. When various inconveniences such as the occurrence of color failure occur, gradation correction and color correction are performed manually, and the correction conditions are stored in the internal memory.

  When the above-described verification operation is completed and a photographic print is output, a main scan, that is, a film image reading process at a high resolution via the film scanner 1a is performed. Various image processes are executed based on the correction conditions described above, and then the print data converted by the print data converter 7e is output to the photo printer 6.

  The above-described pre-scan operation is an operation provided to reduce the number of pixels subject to image processing performed during the above-described verification operation and increase the speed, and is read for an image captured by a digital still camera. The verification operation is performed on the thumbnail image attached to the image file, and when there is no thumbnail image, the verification operation is performed on the thumbnail image generated by thinning the input image.

  Hereinafter, the tone correction processing executed in the above-described verification operation will be described in detail based on the flowchart shown in FIG. The image data input from the image data input unit 1 and stored in the image data storage unit 2 is the image processing memory 7d in the RGB dot sequential RAW format or the RGB plane sequential RAW format in the order of the read frame image unit. (S1), the area dividing means 50a detects the main subject area as the first area (S2), and the other background area is divided as the second area (S3).

  The main subject divided as the first region is a person, animal or other central element captured in the frame image, and various known algorithms are used as the main subject detection algorithm by the region dividing means 50a. Can be realized. For example, when the main subject is a person, the outline based on the density edge and color edge extracted from the original image is learned by using the facial area outline, the arrangement pattern of eyes, nose, mouth, ears, etc. as a teacher signal. When a pattern is recognized as a face area by the network, the area can be divided as the first area.

  It is also possible to detect a human face area from the original image and divide the area including the detected face area as the first area. In order to detect the face area, the skin color and the lip color are set. Various known methods such as a color detection method for extracting and extracting a person, a pattern recognition method for detecting a contour shape of a skin color region, and the like can be used. For example, color feature values such as brightness, hue, saturation, chromaticity, (BR), and (GR) are obtained from the BGR value of each pixel, and the feature values of these colors are within a predetermined range. If it is, the target pixel is determined to be a skin color, edge extraction is performed on an image composed of pixels determined to be a skin color, and a predetermined size and a long axis / short axis (la / lb) are set. The degree of matching between a plurality of ellipse or circular face templates with different ratios (the face template is binarized depending on whether it is an ellipse or a contour of a circle) and the extracted edge image is determined in advance. Disclosed in Japanese Patent Application Laid-Open No. 08-63597 that determines that the face candidate area is equal to or greater than the threshold value and outputs the pair of diagonal coordinates P1 and P2 as a rectangular area having the minimum area including the area. Such as It is also possible.

  Further, it is possible to divide the region based on the correlation of the pixel data of the original image. For example, the original image is divided into a plurality of small areas and a histogram is generated for each divided small area. Areas with similar histograms are recognized as highly correlated areas and areas with significantly different histograms are recognized as weakly correlated areas. It is also possible to separate a strongly correlated region from a weakly correlated region.

  Furthermore, as shown in FIG. 5A, the diagonal coordinates P1, P2 of the rectangular area having the minimum area including the face area are designated by the pointing device such as a mouse from the original image data displayed on the monitor 3. Thus, the rectangular area may be divided into other areas as the first area.

  In this way, the area dividing means 50a generates boundary data between the first area and the second area from the original image, as shown in FIG. 5B, or the first area and the second area. Are stored separately in different layers of the image processing memory 7d (S3).

  The area information generating unit 50c calculates and derives area information indicating image characteristics of the first area and the second area, for example, average density for each area, maximum and minimum density for each area, density histogram for each area, and the like. (S4, S5). These can be calculated for each color such as RGB components, or can be calculated by converting RGB data into luminance data.

  Next, it is determined whether or not an appropriate LUT corresponding to each of the first area including the main subject and the second background image area exists in the LUT storage means 50e (S6), and an appropriate LUT is determined. Is stored, the LUT selection means 50d selects the LUT corresponding to each area (S7, S8). Here, the LUT data corresponding to each area is selected based on the area information of each area obtained by the area information generating means 50b. For example, if the maximum and minimum density ranges are narrow, the LUT is selected so that the gradation range is widened, such that the gradation range is widened. It is. The optimum LUT is selected based on the image characteristics of a single or a combination of image characteristics such as the average density of each area, the maximum density and minimum density of each area, and the density histogram of each area.

  When the image area indicates special area information and an appropriate LUT is not stored in the LUT storage unit 50e, or when a plurality of LUTs are not stored in the LUT data storage unit 50e in advance (S6), An LUT is generated by the LUT data generation means 50f based on the region characteristics (S9, S10). The LUT data created at this time is stored one by one in the LUT data storage unit 50f together with the corresponding area information, so that it can be effectively reused when a similar image is corrected again.

  The LUT data generation unit 50f generates new LUT data so that a preferable gradation is expressed based on a density histogram of the region characteristics. This is because the density histogram is an accumulation of individual pixel data, and is area information including the most image information in each area, so that it is easy to achieve the target gradation. However, the present invention is not limited to this, and it may be created based on the above-mentioned indexes such as the average density of each area, the maximum density and the minimum density of each area.

  The density histogram, which is one of the area characteristics, represents the relationship between the density (gradation) and the number of pixels (frequency) indicating the density based on the pixel data constituting the divided area. It is difficult to generate optimal LUT data for each divided image area. In addition, newly generating LUT data for each correction process may be a factor that hinders the speedup of the correction process inherent in the LUT process. For this reason, LUT data corresponding to a plurality of typical reference density histograms is stored in advance in the LUT storage means 50e, and the actual density histogram is superimposed on the reference density histogram by the LUT selection means 50d. It is preferable that the LUT data corresponding to the closest reference density histogram is selected.

  The LUT processing unit 50d is partitioned by the boundary data generated by the region dividing unit 50a based on the optimum LUT data selected by the LUT selection unit 50d or generated by the LUT data generation unit 50f. The gradation conversion process is executed by executing the conversion process for each area separately stored in the area or layer (S11, S12). An example of the density histogram of the image thus converted is shown in FIG.

  The image converted by the LUT processing unit 50d is developed on the same layer of the image processing memory 7d (S13), and smoothed by the smoothing processing unit 50g so that the boundary region is smoothly expressed (S14). . At this time, the boundary pixel group on the second area side which is the background area among the boundary pixel groups of the first area and the second area so that the gradation of the boundary part of the main subject is not impaired. On the other hand, conversion processing is performed smoothly with the boundary of the first region, that is, with a gentle gradation gradient.

  The image subjected to the gradation correction processing by the gradation correction processing means 50 is then sharpened by the sharpening processing means 52. The sharpening process is performed, for example, by selectively extracting and smoothing a region where graininess or the like is conspicuous by Laplacian filter processing.

  Further, color correction is performed by the color correction means 51. As color correction processing, for example, based on Evans's theory that an average outdoor subject becomes close to gray when the color of the entire negative is mixed, it passes through the negative film when there is a bias in color. In the process to adjust each RGB exposure amount so that the RGB integrated light is reproduced in gray on the photographic paper, the average value of the input image data is calculated and derived for each RGB of each pixel, and each RGB average This is a process of adjusting the values so as to become predetermined values corresponding to gray.

  The correction conditions made during such verification work are stored, correction processing based on the correction conditions is executed on the high-resolution image obtained during the main scan, and the data after image processing is processed by the print data converter 7e. It is converted into print data and output.

  Hereinafter, another embodiment will be described. In the above-described embodiment, the case where there is one first region including the main subject has been described. However, there may be a plurality of main subjects, and the first region may be divided into a plurality of portions. It may be configured to be selected or generated.

  When the region is divided based on the correlation of the pixel data of the original image, as shown in FIG. 6A, for example, the original image is divided into a plurality of small regions and a histogram is generated for each divided small region. It is also possible to recognize a region where the histograms are similar to each other as a strongly correlated region, a region where the histograms are significantly different from each other as a weakly correlated region, and separate a strongly correlated region from a weakly correlated region. The result of the LUT process in this case is shown in FIG.

External view of photographic image processing device Functional block diagram of the image data processing unit Functional block diagram of the image processing unit Flow chart showing the LUT processing procedure It is explanatory drawing which shows a LUT process sequence, (a) is what the main subject was detected from the original image, (b) is what divided subject image data and background image data, (c) is subject image data and background It is an image data density histogram, and is an explanatory view before and after the LUT processing. FIG. 9D is a composite of the subject image data and background image data subjected to the LUT processing. FIG. 10 is an explanatory diagram of a density histogram in another embodiment, (a) is an explanatory diagram of a density histogram of an original image divided into small areas, and (b) is an explanatory diagram before and after LUT processing of the divided density histogram.

Explanation of symbols

1: Image data input unit 1a: Film scanner 1b: Media driver 2: Image data storage unit 3: Monitor 4: Operation input unit 5: Image processing unit 6: Photo printer 7: System controller 50: Tone correction processing means 50a: Division means 50b: area information generation means 50c: LUT processing means 50d: LUT data selection means 50e: LUT data storage means 50f: LUT data generation means 50g: smoothing processing means 51: color correction means 52: sharpening processing means

Claims (13)

  A photographic image processing method for generating an output image by correcting an input original image, wherein the original image is divided into a first region including a main subject and a second region other than that, A photographic image processing method comprising: an LUT processing step of performing correction processing using different LUTs for the first area and the second area, respectively.   The photographic image processing method according to claim 1, wherein in the LUT processing step, correction processing is performed based on an LUT corresponding to an image characteristic of each divided region.   The photographic image processing method according to claim 1, wherein the LUT is selected from a plurality of preset LUTs.   The photographic image processing method according to claim 1, wherein the LUT is generated based on image characteristics of each divided region.   5. The photographic image processing method according to claim 1, wherein the region dividing step divides the region based on a correlation of pixel data of an original image.   5. The photographic image processing method according to claim 1, wherein the area dividing step divides an area detected from an original image by pattern recognition using a neural network as the first area.   5. The photographic image processing method according to claim 1, wherein in the region dividing step, a human face region is detected from an original image, and a region including the detected face region is divided as the first region.   The photographic image processing method according to any one of claims 1 to 7, further comprising a smoothing processing step of performing a smoothing processing on a boundary region between the first region and the second region after the LUT processing step.   A photographic image processing apparatus that corrects an input original image to generate an output image, an area dividing unit that divides the original image into a first area including a main subject and a second area other than the first area; A photographic image processing apparatus comprising LUT processing means for performing correction processing using different LUTs for the first area and the second area, respectively.   The photographic image processing apparatus according to claim 9, wherein the LUT processing unit performs correction processing based on an LUT corresponding to an image characteristic of each divided region.   The photographic image processing apparatus according to claim 9 or 10, wherein the area dividing unit divides an area detected from an original image by pattern recognition using a neural network as the first area.   The photographic image processing apparatus according to claim 9 or 10, wherein the area dividing unit detects a human face area from an original image, and divides an area including the detected face area as the first area.   The photographic image processing apparatus according to claim 9, further comprising a smoothing processing unit configured to perform a smoothing process on a boundary region between the first region and the second region among images processed by the LUT processing unit.