JP2010186401A - Apparatus, method, and program for processing image and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To generate a binary image with little degradation, for which a background part and an object are separated, from a gray image by a smaller calculation amount. <P>SOLUTION: The image processing apparatus 1 includes: a projection value calculation part 102 for generating the distribution of a projection value indicating the change degree of a density in the gray image; an area division part 103 for dividing the gray image into a plurality of partial areas so that the area of a smaller density change is included in a larger partial area and the area of a larger density change is included in a smaller partial area on the basis of the generated distribution of the projection value; a binarization threshold calculation part 104 for calculating a threshold for binarization in each partial area; a threshold image generation part 105 for interpolating the threshold in each partial area and calculating the threshold for binarization in each pixel of the gray image; and a binarization determination part 106 for binarizing the gray image on the basis of the threshold in each pixel and generating the binary image. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an image processing apparatus, an image processing method, a program, and a recording medium that convert a grayscale image into a binary image consisting of only white and black.

Conventionally, there is a method of extracting objects such as a signboard and a sign from a grayscale image in which a complex landscape is projected by binarizing the grayscale image.
In the techniques described in Patent Documents 1 and 2, in order to obtain a good binary image from which an object can be extracted from a grayscale image whose brightness changes continuously for each portion, the grayscale image is a portion having a predetermined size. A divided image divided equally into regions is generated, and binarization processing is performed for each partial region.
In the technique described in Patent Document 3, a grayscale image is divided into partial images having a plurality of types of sizes, and the appearance distribution of density values in each partial region is obtained. Then, this appearance distribution is approximated by the sum of two single-peak density distributions, and an evaluation amount indicating the degree of appropriateness of the partial region is calculated from the approximation result. Based on this evaluation amount, the partial region is calculated. Select the size. Then, binarization processing is performed for each partial area.

JP 59-114687 A JP 61-194580 A Japanese Patent Laid-Open No. 10-255036

However, in the techniques described in Patent Documents 1 and 2, since it is necessary to match the size of the partial area with the size of the target in the grayscale image, when the size of the target is unpredictable, There is a problem that a binary image in which a background portion and an object are separated is not always obtained.
Moreover, in the technique described in Patent Document 3, it is necessary to generate a plurality of divided images and calculate an evaluation value, which requires a large amount of calculation. In addition, since the size of the partial area is selected from the sizes set in advance, the size of the partial area cannot be adaptively changed and the partial area cannot be optimized. There is. In addition, it is difficult to select an appropriate partial area size for a distribution in which the bimodality is not clear because it follows a simple density distribution, and the background image and the target object are separated from each other and the binary image is less deteriorated. There is a problem that it cannot always be generated.

  The present invention has been made in view of the above points, and an object of the present invention is to perform image processing capable of generating, from a grayscale image, a binary image with less deterioration, in which a background portion and an object are separated, with a smaller amount of calculation. To provide an apparatus, an image processing method, a program, and a storage medium.

  The present invention has been made to solve the above problems, and one aspect of the present invention is a projection value calculation unit that generates a distribution of projection values representing a degree of change in density in a grayscale image, and the projection value calculation. Based on the projection value distribution generated by the image forming unit, the grayscale image is divided into a plurality of partial regions so that a region with a smaller density change is included in a larger partial region, and a region with a larger density change is included in a smaller partial region. An area dividing unit that divides into two, a partial region threshold calculating unit that calculates a threshold value for binarization in each partial region divided by the region dividing unit, and the partial region calculated by the partial region threshold calculating unit A pixel threshold value calculation unit that calculates a threshold value for binarization in each pixel of the grayscale image by interpolating the threshold value, and the lightness and darkness based on the threshold value in each pixel calculated by the pixel threshold value calculation unit An image processing apparatus characterized by comprising: a binary image generating unit which generates a binary image by binarizing the image.

  Further, according to an aspect of the present invention, in the image processing apparatus, the projection value calculation unit calculates the projection value corresponding to a scanning line in each of a horizontal direction and a vertical direction in the grayscale image, and the region division The horizontal portion divides the gray image in the horizontal direction so that the product of the horizontal length in each partial region and the sum of the projection values for each scanning line in the horizontal direction in the partial region is equal. The grayscale image is divided in the vertical direction so that the product of the length in the vertical direction and the sum of the projection values for each scanning line in the vertical direction in the partial area are equal.

  Further, according to one aspect of the present invention, in the image processing apparatus, the projection value calculation unit scans the grayscale image on a plurality of scanning lines, and the scanning line having a large density change takes a larger value, The projection value in the scanning line is calculated such that the scanning line having a smaller change takes a smaller value.

  Further, according to one aspect of the present invention, in the above image processing device, the projection value calculation unit includes, for each pixel on the scanning line, a product of a density value of the pixel and a density value of an adjacent pixel, The product of the density value of the pixel and the value obtained by subtracting the density value of the adjacent pixel from the constant, the product of the value obtained by subtracting the density value of the pixel from the constant and the density value of the adjacent pixel, and the density value of the pixel from the constant The projection value in the scanning line is calculated from the product of the value obtained by subtracting the value of the value and the value obtained by subtracting the density value of the adjacent pixel from the constant.

  Further, according to one aspect of the present invention, the projection value calculation unit generates a projection value distribution that represents a degree of change in density in the grayscale image, and the area division unit generates the projection value calculation unit. Based on the projection value distribution, an area that divides the gray image into a plurality of partial areas so that a smaller density change area is included in a larger partial area, and a higher density change area is included in a smaller partial area. A division step, a partial region threshold calculation unit calculates a threshold for binarization in each partial region divided by the region division unit, and a pixel threshold calculation unit calculates the partial region threshold calculation A pixel threshold calculation step of interpolating the threshold value in each partial region calculated by the unit to calculate a threshold value for binarization in each pixel of the grayscale image, and a binary image generation unit, And the binary image generating step of generating a binary image by binarizing the grayscale image based on the threshold value of each pixel element threshold calculating unit is calculated, an image processing method characterized by having a.

  According to another aspect of the present invention, in the image processing method, the projection value calculating step calculates the projection value corresponding to a scanning line in each of a horizontal direction and a vertical direction in the grayscale image, and The step divides the gray image in the horizontal direction so that the product of the horizontal length in each partial area and the sum of the projection values for each scanning line in the horizontal direction in the partial area is equal. The grayscale image is divided in the vertical direction so that the product of the length in the vertical direction and the sum of the projection values for each scanning line in the vertical direction in the partial area are equal.

  According to another aspect of the present invention, in the image processing method described above, the projection value calculating step scans the grayscale image on a plurality of scanning lines, and the scanning line having a larger density change takes a larger value. The projection value in the scanning line is calculated such that the scanning line having a smaller change takes a smaller value.

  Further, according to one aspect of the present invention, in the image processing method, the projection value calculating step includes, for each pixel on the scan line, a product of a density value of the pixel and a density value of an adjacent pixel, The product of the density value of the pixel and the value obtained by subtracting the density value of the adjacent pixel from the constant, the product of the value obtained by subtracting the density value of the pixel from the constant and the density value of the adjacent pixel, and the density value of the pixel from the constant The projection value in the scanning line is calculated from the product of the value obtained by subtracting the value of the value and the value obtained by subtracting the density value of the adjacent pixel from the constant.

  Further, according to one aspect of the present invention, based on the step of generating a projection value distribution indicating the degree of change in density in a grayscale image and the generated distribution of the projection value, an area with a smaller change in density is divided into a larger partial area. A step of dividing the grayscale image into a plurality of partial regions so as to be included in a smaller partial region as a region including a larger change in density, and calculating a threshold value for binarization in each divided partial region; Interpolating the threshold value in each calculated partial region to calculate a threshold value for binarization in each pixel of the grayscale image, and binarizing the grayscale image based on the calculated threshold value in each pixel And a program for causing a computer to execute a step of generating a binary image.

  In addition, according to one aspect of the present invention, based on the step of generating a projection value distribution indicating the degree of change in density in a grayscale image and the generated distribution of the projection value, an area with a smaller change in density has a larger partial area. A step of dividing the grayscale image into a plurality of partial regions so as to be included in a smaller partial region as a region including a larger change in density, and calculating a threshold value for binarization in each divided partial region; Interpolating the threshold value in each calculated partial area to calculate a threshold value for binarization in each pixel of the grayscale image, and binarizing the grayscale image based on the calculated threshold value in each pixel A computer-readable recording medium on which a program for causing a computer to execute a step of generating a binary image by converting the program is recorded.

  According to the present invention, since the partial region is determined based on the distribution of the projection value representing the degree of change in density, the background portion is suppressed without setting the size of the object in the image processing device in advance. A partial region where the object can be extracted can be determined dynamically corresponding to the size of the object. Further, since it is not necessary to generate a plurality of divided images, the size of the partial area can be determined with a smaller calculation amount. As a result, a binary image with less deterioration in which the background portion and the object are separated can be generated with a smaller amount of calculation, regardless of the size of the object.

It is the schematic which shows the input grayscale image by one Embodiment of this invention. It is a block diagram which shows the function structure of the image processing apparatus by this embodiment. It is a flowchart which shows the procedure of the binarization process by this embodiment. It is the schematic for demonstrating the projection value calculation process by this embodiment. It is the schematic for demonstrating the area | region division process by this embodiment. It is the schematic for demonstrating the binarization threshold value calculation process by this embodiment. It is the schematic for demonstrating the threshold value image generation process by this embodiment.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 1 is a schematic diagram illustrating an input gray image according to an embodiment of the present invention. The terms are defined as follows with reference to FIG.
The density value is a value from 0.0 to 1.0 representing the density of the pixel, and is a relative value where black is 1.0 and white is 0.0.
Reference numeral 200 denotes an input grayscale image. The input grayscale image is a grayscale image to be binarized, for example, an image in which a complex landscape is captured. A grayscale image is an image having density values ranging from 0.0 to 1.0. In the input grayscale image 200 shown in the figure, the hatched portion is the background portion B where the landscape is captured, and the white portion is the object T such as a signboard or a sign.
The binary image is an image obtained by binarizing the density value of the input grayscale image 200 using only white (0.0) and black (1.0).
Further, in the image data, an XY coordinate system is defined with the horizontal direction of the rectangular image as the X-axis direction and the vertical direction as the Y-axis direction.

FIG. 2 is a block diagram illustrating a functional configuration of the image processing apparatus 1 according to the present embodiment.
When the input grayscale image 200 is input, the image processing apparatus 1 is an apparatus that converts the input grayscale image 200 into a binary image 300 and outputs the binary image 300. The image processing apparatus 1 includes a preprocessing unit 101, a projection value calculation unit 102, and a region division. Unit 103, binarization threshold value calculation unit 104, threshold image generation unit 105, and binarization determination unit 106.

The preprocessing unit 101 receives the input grayscale image 200 as input, and outputs to the projection value calculation unit 102 a grayscale image that has been subjected to preprocessing such as size change processing, noise removal processing, and smoothing processing.
The projection value calculation unit 102 receives a grayscale image as input and generates a distribution of projection values in the X-axis direction and Y-axis direction of the input grayscale image. The projection value is a value representing the degree of change in density. The projection value takes a small value in a region where the change in density is small, and the projection value takes a large value in a region where the change in density is large. Then, the projection value calculation unit 102 outputs the generated projection value distribution and the input grayscale image to the region division unit 103.

The region dividing unit 103 receives the projection value distribution and the grayscale image as input, and divides the input grayscale image into a plurality of partial regions based on the input projection value distribution. Then, the region dividing unit 103 outputs the input gray image and the position of the divided partial region to the binarized threshold value calculating unit 104. The position of the partial area is a value representing the range of each partial area with XY coordinate values.
A binarization threshold value calculation unit 104 (partial region threshold value calculation unit) receives the position of a partial region and a grayscale image as input, and calculates a binarization threshold value for binarizing the grayscale image in each partial region. Then, the calculated binarization threshold value of each partial region, the position of each partial region, and the input gray image are output to the threshold image generation unit 105.
The threshold image generation unit 105 (pixel threshold calculation unit) receives the binarization threshold value of each partial region, the position of each partial region, and the grayscale image, and interpolates the binarization threshold value of each input partial region, A binarization threshold value in each pixel for binarizing the grayscale image is calculated, and a threshold image is generated. The threshold image is an image having a pixel value as a binarization threshold. Then, the threshold image generation unit 105 outputs the generated threshold image and the input grayscale image to the binarization determination unit 106.
The binarization determination unit 106 (binary image generation unit) receives the threshold image and the gray image, compares each pixel of the input threshold image and the gray image, generates a binary image 300, and generates the generated 2 A value image 300 is output. That is, the binarization determination unit 106 binarizes the grayscale image based on the binarization threshold value in each pixel and generates a binary image 300.

Next, a detailed procedure of the binarization process by the image processing apparatus 1 will be described with reference to FIG. FIG. 3 is a flowchart showing the procedure of the binarization process according to the present embodiment.
[Step S1: Preprocessing]
First, in step S <b> 1, the preprocessing unit 101 generates a grayscale image obtained by performing preprocessing such as a size change process, a noise removal process, and a smoothing process on the input grayscale image 200. Specifically, in the size changing process, the preprocessing unit 101 enlarges or reduces the input grayscale image 200 so that the size becomes N _x × N _y pixels (N _x and N _y are positive integers). Next, in the noise removal process, for example, the preprocessing unit 101 acquires, for each pixel, the density values of the pixel and eight neighboring pixels (3 × 3 region), and arranges the density values in ascending order. Then, noise removal is performed using an existing method, such as setting the median value at that time as the density value of the pixel. In the smoothing process, the preprocessing unit 101 performs smoothing using an existing method such as averaging the density values of pixels in a predetermined range.

[Step S2: Projection Value Calculation Processing]
Next, in step S <b> 2, the projection value calculation unit 102 generates a projection value distribution in the X-axis direction and the Y-axis direction for the grayscale image preprocessed by the preprocessing unit 101. FIG. 4 is a schematic diagram for explaining the projection value calculation processing according to the present embodiment. In this figure, the array of pixels with coordinate values from (1, j) to (N _x , j) (where (1 ≦ j ≦ N _Y )) is the scanning line SY for the Y coordinate value j, (i, 1) to (i, N _Y ) (where (1 ≦ i ≦ N _x )) is a scanning line SX with respect to the X coordinate value i. The projection value calculation unit 102 scans the grayscale image with a plurality of scans. A projection value is calculated by scanning on a line, where a pixel adjacent to a certain pixel in the arrow direction (positive direction) of the scanning line SX or SY shown in the figure is defined as an adjacent pixel of the pixel. Hereinafter, the constant l is 1.0.

First, the projection value calculation unit 102 calculates the projection value P _x (i) for all X coordinate values i in the grayscale image. Specifically, first, the projection value calculation unit 102 calculates, for the pixels on the scanning line SX with respect to the X coordinate value i, the sum a ap _i of the product a of the density value of the pixel and the density value of the adjacent pixel as follows. Calculated according to equation (1). However, g (i, j) is a density value of the pixel R (i, j) whose XY coordinate value is (i, j) in the grayscale image.

Next, the projection value calculation unit 102 calculates, for a pixel on the scanning line SX with respect to the X coordinate value i, a sum b bp _i of the product b of the density value of the pixel and (l− (density value of adjacent pixel)). It calculates by following Formula (2).

Next, the projection value calculation unit 102 calculates, for the pixels on the scanning line SX with respect to the X coordinate value i, the sum cp _i of the product c of (1−density value of the pixel) and the density value of the adjacent pixel as follows. Calculated according to equation (3).

Next, the projection value calculation unit 102 calculates the product e of (l−density value of the pixel) and (l− (density value of the adjacent pixel)) for the pixel on the scanning line SX with respect to the X coordinate value i. the sum ep _i is calculated by the following equation (4).

Then, the projection value calculating section 102 calculates the projection values _P x (i) by the following equation (5). The projection value P _x (i) takes a small value on the scanning line SX where the density is constant, and takes a large value on the scanning line SX where the density changes frequently.

Next, the projection value calculation unit 102 calculates the projection value P _Y (j) for all the Y coordinate values j in the grayscale image. Specifically, first, the projection value calculation unit 102 calculates the sum ap _{j of a} by the following equation (6) for the pixels on the scanning line SY with respect to the Y coordinate value j.

Next, the projection value calculation unit 102 calculates the total sum bp _j of b by the following equation (7) for the pixels on the scanning line SY with respect to the Y coordinate value j.

Next, the projection value calculation unit 102 calculates the total cp _j of c by the following equation (8) for the pixels on the scanning line SY with respect to the Y coordinate value j.

Next, the projection value calculation unit 102 calculates the total sum ep _j of e for the pixels on the scanning line SY with respect to the Y coordinate value _j by the following equation (9).

Then, the projection value calculation unit 102 calculates the projection value P _Y (j) by the following equation (10). The projection value P _Y (j) takes a small value on the scanning line SY where the density is constant, and takes a large value on the scanning line SY where the density changes frequently.

Then, the projection value calculating unit 102, based on the projection values P _{x (i)} generating a distribution 500 of projection values in the X-axis direction shown in FIG. 4, shown in FIG. 4 based on the projection values P _{Y (j)} A projection value distribution 400 in the Y-axis direction is generated.

  The projection value calculation unit 102 calculates the projection value using the product of the density values of adjacent pixels without using the absolute value of the density value. For this reason, the projection value shows a value corresponding to a change in density stably with respect to a change in brightness of the entire image. That is, the projection value takes a small value on a scanning line with a small change in density and takes a large value on a scanning line with a large change in density. Even in a grayscale image in which the bimodality does not appear clearly in the density distribution, the projection value takes a large value in a region where both the object T and the background part B exist, and only the background part B or the object T It takes a small value in the area where only exists.

In this embodiment, the projection value calculation unit 102 calculates the projection value P _x (i) for all the X coordinate values i. However, for example, the projection value P only for the X coordinate value i that is an even number (or an odd number). _The X coordinate value i for calculating the projection value P _x (i) may be limited, for example, by calculating _x (i). Similarly, for example, even (or odd) may limit the Y coordinate value j for calculating the projection values P _{Y (j)} such as only to calculate the projection values P _{Y (j)} for the Y-coordinate value j is.

[Step S3: Area Division Processing]
Next, in step S <b> 3, the region dividing unit 103 converts the grayscale image into a plurality of parts based on the projection value distribution 500 in the X-axis direction and the projection value distribution 400 in the Y-axis direction generated by the projection value calculation unit 102. Divide into areas. FIG. 5 is a schematic diagram for explaining the region division processing according to the present embodiment. Here, the number Q _{x of} partial areas divided in the X-axis direction and the number Q _{y of} partial areas divided in the Y-axis direction are set in the image processing apparatus 1 in advance. In the example shown in this figure, the area dividing unit 103 divides the grayscale image into 4 × 4 = 16 (A11 to A44) partial areas.

The area dividing unit 103 first divides the grayscale image in the X-axis direction. Specifically, in each partial region, the region dividing unit 103 equals the product of the length in the X-axis direction of the partial region and the sum of the projection values for each scanning line SX in the X-axis direction in the partial region. Thus, the grayscale image is divided in the X-axis direction. TX _k is the product of the length in the X-axis direction in the k-th partial region (1 ≦ k ≦ Q _x ) in the X-axis direction and the sum of the projection values for each scanning line SX in the X-axis direction in the partial region. Is represented by the following equation (11). X _k is the X coordinate value of the right end in the k-th partial region in the X-axis direction. However, _{X 0} is 0.

In the example shown in FIG. 5, the region dividing unit 103 makes TX ₁ in the partial region A11, TX ₂ in the partial region A12, TX ₃ in the partial region A13, and TX ₄ in the partial region A14 equal. determining the X-coordinate values _{X 1} and _{X 2} and _{X 3} for dividing the gray image in the X-axis direction. Specifically, it calculates the _{X k} to satisfy the following equation (12).

Next, the area dividing unit 103 divides the grayscale image in the Y-axis direction. Specifically, in each partial region, the region dividing unit 103 equals the product of the length in the Y-axis direction of the partial region and the sum of the projection values for each scanning line SY in the Y-axis direction in the partial region. Thus, the grayscale image is divided in the Y-axis direction. TY _k is the product of the length in the Y-axis direction in the k-th partial region (1 ≦ k ≦ Q _y ) in the Y-axis direction and the sum of the projection values for each scanning line SY in the Y-axis direction in the partial region. Is represented by the following equation (13). Y _k is the Y coordinate value of the lower end in the k-th partial region in the Y-axis direction. However, _{Y 0} is 0.

In the example shown in FIG. 5, the region dividing unit 103 makes TY ₁ in the partial region A11, TY ₂ in the partial region A21, TY ₃ in the partial region A31, and TY ₄ in the partial region A41 equal. Then, Y coordinate values Y ₁ , Y ₂ and Y ₃ for dividing the grayscale image in the Y-axis direction are determined. Specifically, to calculate a _{Y k} satisfying the following equation (14).

  In the grayscale image, the region including the object T and the background portion B has a large change in density, and thus has a large projection value. On the other hand, the region including only the object T or the background portion B has a small projected value because the change in density is small. Therefore, the grayscale image so that the region including the object T and the background portion B is included in the partial region having a relatively small area, and the region including only the object T or the background portion B is included in the partial region having a relatively large area. Can be divided.

[Step S4: Binarization threshold value calculation process]
Next, in step S4, the binarization threshold value calculation unit 104 calculates a binarization threshold value for binarization in each divided region divided by the region division unit 103. FIG. 6 is a schematic diagram for explaining the binarization threshold value calculation process according to the present embodiment. As shown in the figure, the binarization threshold values of the partial areas A11 to A44 are θ11 to θ44. In the present embodiment, the binarization threshold value calculation unit 104 calculates a binarization threshold value using a discriminant analysis method. Specifically, first, the binarization threshold value calculation unit 104 obtains a frequency distribution of density values of each pixel in the partial area. Then, the binarization threshold value calculation unit 104 divides a pixel having a density value larger than a certain density value θ into a black class and a pixel having a density value equal to or lower than the density value θ into a white class. A high density value θ (maximum variance between classes) is set as a binarization threshold value of the partial area.

[Step S5: Threshold Image Generation Processing]
Next, in step S <b> 5, the threshold image generation unit 105 generates a threshold image based on the binarization threshold value of each partial region calculated by the binarization threshold value calculation unit 104. A threshold image is an image in which the value of each pixel is a binarization threshold for binarizing a grayscale image. FIG. 7 is a schematic diagram for explaining threshold image generation processing according to the present embodiment. First, the threshold image generation unit 105 sets the value of the pixel located at the center of each partial area in the threshold image as the binarization threshold value of each partial area. However, when the length of the partial region in the X-axis direction or the Y-axis direction is an even number, the value obtained by discarding the decimal point of the center coordinate value is used as the coordinate of the pixel located at the center. Next, the threshold image generation unit 105 binarizes a threshold value T (i, j) of an arbitrary pixel R (i, j) located within a quadrangle whose apex is a pixel located at the center of four adjacent partial regions. ) Is calculated from the following equation (15) by linear interpolation. However, θ1 is the binarization threshold value of the upper left vertex of the rectangle, θ2 is the binarization threshold value of the lower left vertex of the rectangle, θ3 is the binarization threshold value of the upper right vertex of the rectangle, and θ4 is the rectangle This is the binarization threshold value of the lower right vertex. L1 is the distance in the X-axis direction from the upper left vertex to the upper right vertex of the rectangle, and L2 is the distance in the Y axis direction from the upper left vertex to the lower left vertex of the rectangle. Α is the distance in the X-axis direction from the upper left vertex of the rectangle to the pixel R (i, j), and β is the distance in the Y-axis direction from the upper left vertex of the rectangle to the pixel R (i, j). is there.

[Step S6: Binarization Determination Processing]
Next, in step S <b> 6, the binarization determination unit 106 generates a binary image 300 by comparing the grayscale image with the threshold image generated by the threshold image generation unit 105. Specifically, the binarization determination unit 106 calculates each pixel value B (i, j) of the binary image 300 from the following equation (16). However, G (i, j) is a pixel value of the pixel R (i, j) in the threshold image.

  As described above, according to the present embodiment, since the partial region is determined based on the distribution of the projection value representing the change in density, the size of the target T is not set in the image processing apparatus 1 in advance. It is possible to determine a partial region in which the background portion B is suppressed and the target T can be extracted, dynamically corresponding to the size of the target. Further, since it is not necessary to generate a plurality of divided images, the size of the partial area can be determined with a smaller calculation amount. Thereby, it is possible to generate a binary image 300 with less deterioration by separating the background portion B and the target object T with a smaller calculation amount regardless of the size of the target object T.

Also, a binarization process is performed by recording a program for realizing each step shown in FIG. 3 on a computer-readable recording medium, causing the computer system to read and execute the program recorded on the recording medium. May be performed. Here, the “computer system” may include an OS and hardware such as peripheral devices.
Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.
The “computer-readable recording medium” means a flexible disk, a magneto-optical disk, a ROM, a writable nonvolatile memory such as a flash memory, a portable medium such as a CD-ROM, a hard disk built in a computer system, etc. This is a storage device.

Further, the “computer-readable recording medium” means a volatile memory (for example, DRAM (Dynamic DRAM) in a computer system that becomes a server or a client when a program is transmitted through a network such as the Internet or a communication line such as a telephone line. Random Access Memory)), etc., which hold programs for a certain period of time.
The program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. Here, the “transmission medium” for transmitting the program refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line.
The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, and what is called a difference file (difference program) may be sufficient.

As described above, the embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to the above, and various design changes and the like can be made without departing from the scope of the present invention. It is possible to
For example, in this embodiment, the discriminant analysis method is used to calculate the binarization threshold value. However, if the method is suitable for calculating the binarization threshold value in the local region, for example, the Niblack method or the density distribution average is used. Other methods such as a method using a value or a median value, or a p-tile method may be used.
In this embodiment, linear interpolation is used as the binarization threshold interpolation method, but other methods for interpolating the threshold may be used.

  DESCRIPTION OF SYMBOLS 1 ... Image processing apparatus 101 ... Pre-processing part 102 ... Projection value calculation part 103 ... Area division part 104 ... Binarization threshold value calculation part 105 ... Threshold image generation part 106 ... Binarization determination part

Claims (10)

A projection value calculation unit that generates a distribution of projection values representing the degree of change in density in the grayscale image;
Based on the projection value distribution generated by the projection value calculation unit, the grayscale image is included so that a region with a smaller change in density is included in a larger partial region and a region with a larger change in density is included in a smaller partial region. An area dividing unit that divides into a plurality of partial areas;
A partial region threshold value calculating unit that calculates a threshold value for binarization in each partial region divided by the region dividing unit;
A pixel threshold value calculation unit that calculates a threshold value for binarization in each pixel of the grayscale image by interpolating the threshold value in each partial region calculated by the partial region threshold value calculation unit;
A binary image generation unit that generates a binary image by binarizing the grayscale image based on the threshold value of each pixel calculated by the pixel threshold value calculation unit;
An image processing apparatus comprising: The projection value calculation unit calculates the projection value corresponding to the scanning lines in the horizontal direction and the vertical direction in the grayscale image,
The region dividing unit divides the gray image in the horizontal direction so that the product of the horizontal length in each partial region and the sum of the projection values for each scanning line in the horizontal direction in the partial region are equal. The grayscale image is divided in the vertical direction so that the product of the vertical length in the partial area and the sum of the projection values for each scanning line in the vertical direction in the partial area are equal. The image processing apparatus described.   The projection value calculation unit scans the grayscale image on a plurality of scanning lines, and the scanning line has a larger value as the scanning line having a larger density change and takes a smaller value as the scanning line has a smaller density change. The image processing apparatus according to claim 1, wherein the projection value in a line is calculated.   The projection value calculation unit is a value obtained by subtracting the density value of the adjacent pixel from the product of the density value of the pixel and the density value of the adjacent pixel and the density value of the pixel and a constant in each pixel on the scanning line. , The product of the value obtained by subtracting the density value of the pixel from the constant and the density value of the adjacent pixel, the value obtained by subtracting the density value of the pixel from the constant, and the value obtained by subtracting the density value of the adjacent pixel from the constant The image processing apparatus according to claim 3, wherein the projection value in the scanning line is calculated from the product of A projection value calculating step for generating a distribution of projection values representing a degree of change in density in the grayscale image;
Based on the projection value distribution generated by the projection value calculation unit, the region dividing unit includes a region with a smaller change in density as a larger partial region and a region with a larger change in density as a smaller partial region. An area dividing step for dividing the gray image into a plurality of partial areas;
A partial region threshold value calculating unit that calculates a threshold value for binarization in each partial region divided by the region dividing unit;
A pixel threshold calculation unit that interpolates the threshold in each partial region calculated by the partial region threshold calculation unit and calculates a threshold for binarization in each pixel of the grayscale image;
A binary image generation step of generating a binary image by binarizing the grayscale image based on the threshold value of each pixel calculated by the pixel threshold value calculation unit;
An image processing method comprising: The projection value calculating step calculates the projection value corresponding to the scanning lines in the horizontal direction and the vertical direction in the grayscale image,
The region dividing step divides the gray image in the horizontal direction so that the product of the horizontal length in each partial region and the sum of the projection values for each scanning line in the horizontal direction in the partial region are equal. 6. The gray image is divided in the vertical direction so that the product of the length in the vertical direction in the partial area and the sum of the projection values for each scanning line in the vertical direction in the partial area is equal. The image processing method as described.   The projection value calculating step scans the grayscale image on a plurality of scanning lines, and the scanning line having a larger density change takes a larger value, and the scanning line having a smaller density change takes a smaller value. The image processing method according to claim 5, wherein the projection value in a line is calculated.   In the projection value calculating step, for each pixel on the scanning line, a product of the density value of the pixel and the density value of the adjacent pixel, and a value obtained by subtracting the density value of the adjacent pixel from the density value and constant of the pixel. , The product of the value obtained by subtracting the density value of the pixel from the constant and the density value of the adjacent pixel, the value obtained by subtracting the density value of the pixel from the constant, and the value obtained by subtracting the density value of the adjacent pixel from the constant The image processing method according to claim 7, wherein the projection value in the scanning line is calculated from the product of Generating a distribution of projection values representing the degree of change in density in the grayscale image;
Based on the generated projection value distribution, the grayscale image is divided into a plurality of partial areas so that areas with smaller changes in density are included in larger partial areas and areas with higher changes in density are included in smaller partial areas. And steps to
Calculating a threshold value for binarization in each divided partial area;
Interpolating the threshold value in each calculated partial area to calculate a threshold value for binarization in each pixel of the grayscale image;
Binarizing the grayscale image based on the calculated threshold value for each pixel to generate a binary image;
A program that causes a computer to execute. Generating a distribution of projection values representing the degree of change in density in the grayscale image;
Based on the generated projection value distribution, the grayscale image is divided into a plurality of partial areas so that areas with smaller changes in density are included in larger partial areas and areas with higher changes in density are included in smaller partial areas. And steps to
Calculating a threshold value for binarization in each divided partial area;
Interpolating the threshold value in each calculated partial area to calculate a threshold value for binarization in each pixel of the grayscale image;
Binarizing the grayscale image based on the calculated threshold value for each pixel to generate a binary image;
The computer-readable recording medium which recorded the program for making a computer perform.

Images