JP2007104706A - Image processing apparatus and image processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To automatically optimize block size for block-splitting a processing-object image, and to automatically perform an optimum highlighting of contrast per block. <P>SOLUTION: For example, an image taken with a digital camera is input in a region splitting part through an image input part. The region splitting part, which splits the input image into a plurality of blocks, optimizes sizes of the blocks according to similarity of density histogram of each block and splits them. A contrast highlight part calculates an optimum clip value for each of the split blocks, and uses the clip value to calculate a density transformation curve from the density histogram in the block to which an attentional pixel belongs to optimally perform density transformations of all the pixels. In this way, the contrast-highlighted image is corrected to properties of a printing device by a read correction part before being printed on a paper and output. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an image processing apparatus and an image processing apparatus that perform image processing so that an output image without crushing or jumping is obtained by enhancing the contrast of the image when the image is printed on paper or the like by a printing apparatus or the like. It relates to the processing method.

  Conventionally, a printer having a function of printing an image output from a personal computer such as an ink jet printer or a laser printer on paper, an arbitrary image such as a copying machine is read by an optical reader, and another paper is used. A copying apparatus having a function of printing on a paper has been developed.

  In such an image printing apparatus such as a printer or a copying apparatus, contrast enhancement by histogram equalization or the like is performed in order not to lose fine edge information of the original image.

  There is a local histogram equalization method as a general contrast enhancement method. This method performs contrast enhancement in consideration of local information of an image, and is considered to be effective for a natural image in which local information is important.

  However, this method has a drawback that the processing time becomes enormous because a mapping curve (density value conversion curve), which is an accumulation of density value histograms, is obtained for each pixel.

  In order to avoid this drawback, for example, a paper “Navigating with high-speed local contrast for natural images” published by Naoki Kobayashi (Electronic Information and Communication Society Journal D-II Vol. J77-D-II No. 3pp. 502-509) describes a method of obtaining a mapping curve for each region instead of for each pixel in order to shorten the histogram creation time, thereby speeding up the histogram creation time.

  Next, the outline of the method will be described below.

  (1) A plurality of density conversion curves are designed in advance.

  (2) Assuming that the histograms are concentrated near the average density, a density conversion curve is selected according to the average density (selecting a density conversion curve that emphasizes the contrast near the average density).

  (3) When the selected density conversion curves are different in adjacent areas, linear interpolation of density values is performed.

  By the way, in the above paper, it is described that a density conversion curve is created for each pixel, but a method for creating a density conversion curve for each block as described below is more general.

1) Divide the input image into blocks of uniform size determined empirically.

2) The following processes (1) to (3) are performed for each block.

  (1) A density histogram is created (a block is used as a reference area).

  (2) Clip the density histogram with clip values determined empirically to obtain a density histogram after clipping.

  (3) A cumulative histogram is created by accumulating the density histogram after clipping. 3) Perform density conversion of each pixel in the block using the accumulated histogram as a density conversion curve. However, when the density conversion curve of the block to which the pixel of interest belongs differs from the density conversion curve of the adjacent block, linear interpolation of density values as follows is performed.

  (1) The density value of the target pixel is density-converted using the density conversion curve created in the block B1 in which this pixel exists, and the converted density value g1 is obtained.

  (2) The density value of the target pixel is density-converted using density conversion curves selected in each of the three blocks B2, B3, and B4 adjacent to this pixel, and converted density values g2, g3, and g4 are obtained.

  (3) The density value g (x, y) after linear interpolation is calculated by the expression 1 (the density values g1, g2, g3, and g4 after conversion are noticed from the centers of the four blocks B1, B2, B3, and B4). Weighted according to the distance to the pixel).

g (x, y) = {(J−j) / J} {(I−i) g1 / I + ig2 / I} + j / J {(I−i) g3 / I + ig4 / I} (1)
However, for the definition of each variable in this equation, see the description of the same equation used in the examples.
"High-speed local contrast enhancement for natural images" (The Institute of Electronics, Information and Communication Engineers Journal D-II Vol. J77-D-II No. 3pp. 502-509)

  However, in the above-mentioned paper “High-speed local contrast enhancement for natural images”, a region having a low density value and a small area (region A) and a region having a high density value and a large area (region) in the target image. B), if the size of the reference region is optimally set according to the region A, only the local contrast may be enhanced in the region B and the overall contrast may be lowered. .

  Conversely, if the size of the reference area is optimally set according to the area B, the size of the area A is very small relative to the size of the reference area, and the density information of the area A contributes greatly when calculating the histogram. As a result, the slope of the low density portion of the density conversion curve becomes small and the contrast may be lowered.

  After all, since the areas for which the mapping curve is obtained are all the same size, there is a problem that the size of the area cannot be optimized for the entire processed image.

  In addition, it is desirable that the clip value, which is a parameter for determining the size of the reference region and the parameter for determining the degree of contrast enhancement, which are parameters for determining the degree of enhancement of the detailed information, is determined according to the characteristics of the local region of the image. Since there is no such determination method, conventionally, a fixed value determined empirically is used. For this reason, it is desired that these parameters can be automatically determined according to local characteristics of the image.

  As the reference area is smaller, the local contrast is enhanced and the global contrast is lowered. Conversely, as the reference area is larger, the global contrast is enhanced and the local contrast is lowered. Furthermore, the degree of emphasis increases as the clip value increases, and the degree of emphasis decreases as the clip value decreases.

  Furthermore, when the target image changes, the optimum parameter also changes, so that trial and error must be repeated each time, which is inefficient.

  Further, since the display device has a sufficiently wide density dynamic range, there is a problem in that over-emphasis of contrast may occur when the density dynamic range of the image display device is narrower than the original image.

  The present invention has been made in order to solve the above-described conventional problems. The object of the present invention is to automatically optimize the block size of the reference region and automatically optimize the contrast enhancement for each region. It is an object of the present invention to provide an image processing apparatus and an image processing method that can be performed and can easily obtain an output image that does not cause a sense of incongruity and that is free of collapse or jumping regardless of the skill level of the operator.

  In order to achieve the above object, the invention according to claim 1 is characterized in that, in an image processing apparatus for outputting an image with enhanced contrast of an input image to an output device, an area dividing means for dividing the input image into a plurality of blocks. A contrast enhancement unit that performs density conversion of the image in the block at a degree of contrast enhancement corresponding to the density characteristic of the pixel for each block obtained by the region dividing unit, and the contrast enhancement unit includes: A clip which is a parameter for calculating the slope of the tangent line near the average density of each block on the output system characteristic curve, which is the characteristic of the output device, and determining the degree of the tangent line and the degree of contrast enhancement determined in advance. The clip value in each block is determined based on the relationship with the value.

According to a second aspect of the present invention, there is provided an image processing apparatus that outputs an image in which contrast of an input image is enhanced to an output device, and is obtained by an area dividing unit that divides the input image into a plurality of blocks and the area dividing unit. Contrast enhancement means for converting the density of the image in the block at a degree of contrast enhancement corresponding to the density characteristic of the pixel for each block, and the contrast enhancement means is an output that is a characteristic of the output device. Calculate the slope of the straight line connecting the two points corresponding to the average soil concentration variance of each block on the system characteristic curve,
A clip value in each block is determined based on a relationship between a predetermined slope of the straight line and a clip value that is a parameter for determining the degree of contrast enhancement.

  The feature of the invention of claim 3 is that the contrast enhancement means divides a block larger than this in accordance with the size of the minimum block for pixels in the area corresponding to the boundary of the block, and The purpose is to perform density conversion using a density curve obtained by linear interpolation of the density curve.

  A feature of the invention of claim 4 is that the contrast enhancing means obtains a difference in average density between all adjacent blocks, and includes a total of four blocks including a block in which the pixel of interest exists and three blocks closest to the pixel of interest. Create block pairs in the vertical, horizontal, and horizontal directions, determine whether the difference in the average density of each block pair is greater than or equal to a preset threshold, and reduce the degree of contrast enhancement according to the number of block pairs greater than or equal to the threshold There is to make it.

  According to a fifth aspect of the present invention, there is provided an image processing method for outputting an image in which contrast of an input image is enhanced to an output device, a process of dividing the input image into a plurality of blocks, and pixels obtained in the obtained blocks A clip value for determining the degree of contrast enhancement of each block is obtained from the process of obtaining the average of the density values of each block, and the obtained average density value of each block and the output system characteristic curve which is a characteristic of the output device. A process of obtaining a density histogram of each block obtained, a process of creating a density conversion curve of each block from the obtained density histogram and clip value of each block, and performing density conversion of all pixels It is in having.

  The feature of the invention of claim 6 is that a density average difference between all the adjacent blocks in the obtained block is obtained, and a total including a block in which the pixel of interest exists and three blocks closest to the pixel of interest. The process of creating block pairs in the vertical and horizontal directions of 4 blocks, the process of determining whether the difference in average density of each block pair is greater than or equal to a preset threshold, and the number of block pairs that are greater than or equal to the threshold by the determination And a step of reducing the clip value that determines the degree of contrast enhancement according to the above.

  The feature of the invention of claim 7 is that, for the pixels in the region corresponding to the boundary of the block, a process of dividing a block larger than this in accordance with the size of the minimum block and linear interpolation of the density curve of each block. And a step of performing density conversion using the obtained density curve.

  The feature of the invention of claim 8 is obtained in the image processing device for outputting an image with enhanced contrast of the input image to the output device by the region dividing means for dividing the input image into a plurality of blocks and the region dividing means. Contrast enhancement means for converting the density of the image in the block at a degree of contrast enhancement corresponding to the density characteristic of the pixel for each block, and the contrast enhancement means is an output that is a characteristic of the output device. The clip value in each block is determined based on the system characteristic curve and the average density of each block.

  According to the above invention, the input image is automatically divided into a plurality of blocks. At that time, it is determined whether or not to further divide the block according to the similarity of the density histogram of the block. If division is necessary, the block is repeatedly divided into smaller blocks. It is automatically divided into optimally sized blocks.

  Thereafter, the degree of contrast enhancement is automatically determined for each obtained block, and density conversion is performed on all pixels. At that time, the average density value of the pixels in each block is obtained, and the degree of contrast enhancement of each block is determined from the average density value of each block and the output system characteristic curve which is a characteristic of the output device such as a printer. The optimum clip value is automatically obtained, and after obtaining the density histogram of each block, the density conversion curve of each block is created from the density histogram of each block and the clip value to perform density conversion of all pixels. Thus, the contrast enhancement of the input image can be optimally performed regardless of the skill level of the operator.

  As described above in detail, according to the first aspect of the present invention, the clip value, which is a parameter for determining the emphasis degree of each block, is automatically obtained for each block using the average density and the output system characteristic curve. Can do.

  According to the second aspect of the present invention, the average density and variance are calculated for each block, and the clip value is determined using the output system characteristic curve. It becomes difficult to be affected by noise.

  According to the third and seventh aspects of the invention, block noise can be prevented and the quality of the output image can be improved.

  According to the fourth and sixth aspects of the invention, over-emphasis of contrast can be prevented, and the naturalness of the output image can be improved.

  According to the invention of claim 5, the optimum clip value that is a parameter for determining the degree of emphasis of each block can be automatically obtained for each block, and this is used to automatically optimize the contrast for each block. Since the emphasis can be made, an output image free from crushing and jumping can be obtained easily and stably regardless of the skill level of the operator.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a configuration of an embodiment of an image processing apparatus of the present invention. Connected to the subsequent stage of the image processing apparatus 1 is a printing apparatus 3 that prints processed images on paper or the like and outputs them.

  The image processing apparatus 1 includes, for example, an image input unit 11 for inputting image data captured by a separate digital camera or the like from the outside, an output system characteristic curve indicating the output system characteristic of the printing apparatus 3, and an output obtained by inversely converting the output system characteristic curve. Using the memory unit 12 that stores the system correction curve, the region dividing unit 13 that divides the region according to the characteristics of the local region of the processing target image (same as the print image), the output system correction curve that is obtained in advance, and the like. The image processing apparatus includes a contrast enhancement unit 14 that performs various processes for performing optimum enhancement of contrast for each region, and a reading correction unit 15 that performs density correction of a print image using the output system characteristic curve.

  Next, the operation of this embodiment will be described. First, in order to obtain the output system characteristic curve and the output system correction curve stored in the memory unit 12 of the image processing apparatus 1, the processing shown in the flowchart of FIG. 2 is performed in advance. In step 201, the density dynamic range of the printing apparatus 3 is quantitatively measured. FIG. 3 is a diagram showing an example of a test pattern. By printing such a test pattern by the printing apparatus 3 and analyzing the printing result, an output system characteristic curve which is a characteristic of the printing apparatus 3 can be obtained in step 202.

  FIG. 4 is a characteristic diagram showing an example of the output system characteristic curve of the printing apparatus 3 obtained in this way. From this input / output characteristic, an output system correction curve as shown in FIG. 5 is obtained so that there is no difference in absolute density information between the original image and the printed image. These output system characteristic curve and output system correction curve are stored in the memory unit 12 in advance.

  Next, the operation of the image processing apparatus 1 will be described with reference to the flowchart of FIG. The image processing apparatus 1 performs a process for contrast enhancement on the print target image so that the fine edge information is not lost due to being buried in the noise of the printing apparatus 3.

  First, in step 601, the image input unit 11 of the image processing apparatus 1 temporarily stores image data (image to be printed) input from outside in an image memory (not shown), and then stores the input image data in a region. The data is output to the dividing unit 13 and the contrast enhancing unit 14.

  Thereafter, the region dividing unit 13 automatically divides the print target image into a plurality of blocks (regions) in step 602, and in step 603, the contrast enhancing unit 14 is a parameter for determining the degree of contrast enhancement. The clip value is automatically determined.

  Here, the local histogram equalization method will be described with reference to the drawings so that the characteristic part of this example can be easily understood. First, as shown in FIG. 7, a reference area 003 having an appropriate size is provided around the target pixel 002 in the processed image 001 for density conversion, and a density histogram 004 of the pixels in the reference area 003 is created. Next, in order to suppress excessive emphasis, a pixel having a frequency exceeding an appropriate clip value 004a is clipped from the obtained histogram, and a histogram 005 is obtained by equally distributing all the density values. Then, the accumulated histogram 006 obtained by accumulating the clipped histogram is normalized to obtain a density conversion curve, and the density conversion of the pixel of interest is performed. However, with this method as it is, it is necessary to calculate a density histogram for all the pixels in the image, so that the processing time becomes enormous.

  Therefore, in step 602, the area dividing unit 13 of the image processing apparatus 1 of the present example divides the processed image into a plurality of blocks 013, 014,... As shown in FIG. By using the curve, the processing time is greatly shortened as in the prior art. In addition, since low density areas and high density areas are localized in the image to be processed, the entire processed image is divided into square blocks of different sizes as shown in FIG. is doing.

  In this block division, since the same density conversion curve is used in the same area, the same area is preferably divided into the same blocks. Therefore, as shown in FIG. 8, (1) the entire processed image is divided into blocks of uniform size (this block is referred to as a “large block”).

  (2) Whether to divide the obtained large block 013 or not is determined by the following procedure. First, in the process of (2-1), the large block 013 is virtually divided into four virtual blocks of uniform size (this size block is referred to as “medium block”), and (2-2) In this process, density histograms 018, 019, 020, and 021 for the four virtual medium blocks 014, 015, 016, and 017 are created.

  Next, in the process of (2-3), these four density histograms 018, 019, 020, 021 are overlapped as shown by 022, and the area S of the overlapping portion 023 and a predetermined threshold T ( Compare empirically). As a result, when the area S is equal to or greater than the threshold T, it is determined that the similarity of the four virtual middle blocks 014, 015, 016, and 017 is high, and these virtual middle blocks are integrated and returned to the large block 013. That is, it is determined that the large block is not further divided. On the other hand, when the area S is smaller than the threshold value T, it is determined that the four virtual medium blocks 014, 015, 016, and 017 are low in similarity and are divided as formal medium blocks.

Thereafter, in the process (2-4), the processes (2-1) to (2-3) described above are performed on all large blocks.

  (3) If it is determined that the middle block obtained in the process (2) above is to be divided by replacing the large block in (2) with the middle block, The medium block is further divided into small blocks.

  In the above description, an example of dividing into blocks of “large”, “medium”, and “small” is shown. However, such processing is performed recursively, and further, “large”, “medium” You may make it divide | segment into any block of "small" "minimal".

  By the way, when the block division described with reference to FIG. 8 is performed, the density conversion curves are different at the boundaries between the blocks, so that density values are discontinuous at the boundaries and block noise occurs.

  Therefore, block noise can be prevented by using proportional distribution of density conversion curves of four surrounding blocks (009 to 012) for the pixel of interest 008 shown in FIG. By proportional distribution, the output value g (x, y) of the target pixel 008 is expressed by the following equation.

g (x, y) = {(J−j) / J} {(I−i) g1 / I + ig2 / I} + j / J {(I−i) g3 / I + ig4 / I} (1)
However, g1 to g4 are density values converted by density conversion curves obtained in four blocks of a block where the target pixel exists and a block closest to the target pixel, and J and I are the target pixels shown in FIG. The lengths of the sides of the quadrangle included, i and j indicate the position of the pixel of interest 008 with reference to one vertex of the quadrangle.

  If the size of the block around the pixel of interest 008 is different, as shown in FIG. 10, the large and medium blocks are virtually divided into blocks according to the size of the small blocks, and the blocks indicated by diagonal lines in the figure are displayed. For pixels in the region corresponding to the boundary, density conversion may be performed using Equation (1) so that block noise does not occur.

  The contrast enhancement unit 14 uses the same clip value in the same region for each block divided by the region division unit 13. Here, the clip value is a parameter that determines the degree of contrast enhancement. The greater the clip value, the greater the enhancement. On the other hand, if the clip value is 0, no contrast enhancement is performed.

This determination method obtains the average density m and density variance (σ ² ) of the pixels in the block, and compares them with an output system characteristic curve (printer characteristics) 100 as shown in FIG. Determine the clip value.

  Specifically, as shown in FIG. 11, the density value of the pixel in the corresponding block is considered to be localized between (m−σ) and (m + σ), and these two points (average density ± dispersion) The larger the slope of the straight line 025 connecting the same), the wider the dynamic range, that is, the larger the number of output gradations relative to the number of input gradations.

  Therefore, in step 603, the contrast emphasizing unit 14 considers that the dynamic range in the block is wide and the loss of information is small, and if the clip value is reduced and the inclination is small, the loss of information is considered large. The optimum clip value is obtained for each block so as to increase the clip value.

  The contrast enhancement unit 14 can also obtain the clip value using only the average density. In this case, the contrast enhancement unit 14 calculates the slope of the tangent line a at the point where the input corresponds to the average density m on the output system characteristic curve P as shown in FIG. A clip value corresponding to the slope of the tangent a is obtained by referring to the relationship with the value.

Further, according to the method of determining the clip value using the average density m and its variance σ ² , it is less affected by the noise of the output system characteristic curve as compared with the case of determining using only the average density m. There is an advantage, and when the determination is made using only the average density m, there is an advantage that the arithmetic processing is reduced as compared with the case where the variance σ ² is also used. Note that the average density m does not have to be one point, and the clip value can be obtained within a negligible error range as long as the point is in the vicinity thereof.

  Thereafter, in step 604, the contrast emphasizing unit 14 creates a density conversion curve from the density histogram in the block to which the pixel of interest belongs by using the above-described area division information and this clip value information, and performs density conversion of all the pixels. Do.

  Here, if the density conversion curves between adjacent blocks are greatly different, that is, if the features between the blocks are greatly different, the clip value is decreased because the contrast is sufficiently large and there is little need for emphasis. Specifically, in a total of four blocks including the block to which the target pixel belongs and three blocks adjacent to the target pixel, every time the difference in average density m between the upper, lower, left, and right blocks is equal to or greater than a preset threshold value, The clip value is decreased by a predetermined amount. Each time the difference in the average density m of a set of blocks exceeds the threshold, the clip value is decreased by ¼, and the clip value is set to 0 if the threshold is exceeded in all sets of upper, lower, left, and right.

  Therefore, the contrast emphasizing unit 14 obtains a difference in average density between all adjacent blocks, and a block pair in the vertical and horizontal directions of a total of four blocks including the block in which the pixel of interest exists and the three blocks closest to the pixel of interest. To determine whether the difference in the average density of each block pair is greater than or equal to a preset threshold, and to reduce the clip value according to the number of block pairs that are greater than or equal to the threshold to suppress contrast enhancement .

  The print target image whose contrast is enhanced by the contrast enhancement unit 14 is sent to the reading correction unit 15. In step 605, the reading correction unit 15 uses the output system correction curve as shown in FIG. 5 stored in the memory unit 12 to the printing apparatus 3 to output absolute density information of the original image and the print image. The density is inversely converted so as to eliminate the difference, and the obtained image data is sent to the printing apparatus 3 to output a print image.

  According to the present embodiment, the image to be printed is automatically divided into a plurality of optimally sized blocks by the area dividing unit 13, and the optimum clip value is automatically determined by the contrast enhancing unit 14 for each block. Accordingly, since the contrast enhancement process is performed for each block, a print image with the optimum contrast enhancement can always be obtained easily and quickly regardless of the skill level of the operator.

  In addition, since the block for obtaining the density conversion curve has an optimal size according to the density change of the processed image, the contrast of the processed image can be optimally enhanced as a whole. In addition, when the average density value between all adjacent blocks exceeds a certain value, the clip value is decreased to suppress the contrast enhancement, so that the contrast overemphasis can be eliminated. Due to these effects, even when the density dynamic range of the printing apparatus 3 is narrower than the dynamic range of the output image, it is possible to eliminate the loss of information and to obtain an output image that is not crushed or jumped easily and stably. Can do.

  In the above embodiments, an apparatus having a function of printing an image output from a personal computer or the like such as an ink jet printer or a laser printer on paper or the like has been described. However, an optical reading device for reading an arbitrary image such as a copying machine. The same effect can be obtained even if the present invention is applied to an apparatus having the apparatus. Further, a scanner or the like is connected to the image processing apparatus, and image data read by the scanner is input from the image input unit 11 and processed, that is, the scanner, the image processing apparatus 1, and the printer (separate or integrated). Even in this configuration, the same effect can be obtained by applying the present invention.

  This is useful for automatically optimizing the block size when the image to be processed is divided into blocks.

1 is a block diagram showing a configuration of an embodiment of an image processing apparatus of the present invention. 3 is a flowchart showing a procedure for acquiring an output system characteristic curve stored in the memory unit shown in FIG. 1. It is the figure which showed the example of a test pattern used when acquiring the said output type | system | group characteristic curve. FIG. 2 is a characteristic diagram illustrating an example of an output system characteristic curve stored in the memory unit illustrated in FIG. 1. FIG. 3 is a characteristic diagram illustrating an example of an output system correction curve stored in the memory unit illustrated in FIG. 1. 3 is a flowchart showing an image processing procedure of the image processing apparatus shown in FIG. 1. It is a figure explaining the method of applying the histogram equalization method to a local area | region, and calculating | requiring the density conversion curve for every pixel. It is a figure explaining the area | region division operation | movement of the area | region division part shown in FIG. It is a figure explaining the example of the interpolation method for preventing the discontinuity of the density value in the boundary of the block of a fixed magnitude | size. It is a figure explaining the other example of the interpolation method for preventing the discontinuity of the density value in the boundary of the block of a different magnitude | size. It is a figure explaining the method by which the contrast emphasis part shown in FIG. 1 calculates | requires a clip value. It is a figure explaining the other method in which the contrast emphasis part shown in FIG. 1 calculates | requires a clip value.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Image processing apparatus 3 Printing apparatus 11 Image input part 12 Memory part 13 Area division part 14 Contrast emphasis part 15 Reading correction part 001,007 Processed image 002,008 Target pixel 003 Reference area 004, 018-021 Density histogram 005 After clipping Histogram 006 Cumulative histogram 009 to 012 4 blocks around the pixel of interest 013 to 017 block 023 Area of overlapping portion 025 Straight line connecting m−σ, m + σ

Claims (8)

In an image processing device that outputs an image with enhanced contrast of an input image to an output device,
Area dividing means for dividing the input image into a plurality of blocks;
Contrast enhancement means for performing density conversion of the image in the block at a degree of contrast enhancement corresponding to the density characteristic of the pixel for each block obtained by the region dividing means,
The contrast enhancement means calculates an inclination of a tangent line in the vicinity of an average density of each block on an output system characteristic curve which is a characteristic of the output device,
An image processing apparatus that determines a clip value in each block based on a relationship between a slope of a tangent line that is determined in advance and a clip value that is a parameter that determines the degree of contrast enhancement. In an image processing device that outputs an image with enhanced contrast of an input image to an output device,
Area dividing means for dividing the input image into a plurality of blocks;
Contrast enhancement means for performing density conversion of the image in the block at a degree of contrast enhancement corresponding to the density characteristic of the pixel for each block obtained by the region dividing means,
The contrast enhancement means calculates an inclination of a straight line connecting two points corresponding to an average density soil dispersion of each block on an output system characteristic curve which is a characteristic of the output device, and a predetermined straight line inclination And a clip value that is a parameter for determining the degree of contrast enhancement, the clip value in each block is determined.   The contrast enhancement means is obtained by linearly interpolating the density curve of each block after dividing a block larger than this in accordance with the size of the minimum block for pixels in the region corresponding to the boundary of the block. 3. The image processing apparatus according to claim 1, wherein density conversion is performed using a density curve. The contrast enhancement means calculates an average density difference between all adjacent blocks,
Create a total of 4 blocks, up / down / left / right block pairs, including the block where the pixel of interest exists and the 3 blocks closest to the pixel of interest,
Determine whether the average density difference of each block pair is greater than or equal to a preset threshold,
The image processing apparatus according to claim 1, wherein the degree of contrast enhancement is reduced according to the number of block pairs equal to or greater than a threshold value. In an image processing method for outputting an image with enhanced contrast of an input image to an output device,
Dividing the input image into a plurality of blocks;
Obtaining an average of the density values of the pixels in each obtained block;
Obtaining a clip value for determining the degree of contrast enhancement of each block from the average of the obtained density values of each block and an output system characteristic curve which is a characteristic of the output device;
Obtaining a density histogram of each of the obtained blocks;
A process of creating a density conversion curve of each block from the obtained density histogram and clip value of each block and performing density conversion of all pixels;
An image processing method comprising: Find the average density difference between all adjacent blocks in the obtained block,
A process of making a block pair in the vertical and horizontal directions of a total of 4 blocks including a block in which the pixel of interest exists and 3 blocks closest to the pixel of interest;
A process of determining whether the difference in average density of each block pair is equal to or greater than a preset threshold;
Reducing the clip value that determines the degree of contrast enhancement according to the number of block pairs that are equal to or greater than the threshold value according to the determination;
The image processing method according to claim 5, further comprising: For pixels in the region corresponding to the boundary of the block, a process of dividing a larger block according to the size of the minimum block;
The process of density conversion using the density curve obtained by linear interpolation of the density curve of each block,
The image processing method according to claim 5, further comprising: In an image processing device that outputs an image with enhanced contrast of an input image to an output device,
Area dividing means for dividing the input image into a plurality of blocks;
Contrast enhancement means for performing density conversion of the image in the block at a degree of contrast enhancement corresponding to the density characteristic of the pixel for each block obtained by the region dividing means,
The contrast enhancement means is based on an output system characteristic curve which is a characteristic of the output device and an average density of each block.
An image processing apparatus for determining a clip value in each block.

Images