KR20160065474A

KR20160065474A - Method and apparatus for image processing using histogram

Info

Publication number: KR20160065474A
Application number: KR1020140169466A
Authority: KR
Inventors: 김지호; 차준호; 정재호; 정직한
Original assignee: 한화테크윈 주식회사
Priority date: 2014-12-01
Filing date: 2014-12-01
Publication date: 2016-06-09
Also published as: KR101981039B1

Abstract

A method of processing an image according to an exemplary embodiment of the present invention includes generating a histogram using pixels included in an image, extracting candidate pixels corresponding to a brightness of a detection target to be detected from the image in the histogram, Calculating a degree of variance representing a degree of variance of the candidate pixels using the in-view coordinates of the candidate pixels; identifying a target pixel of the candidate pixels using the variance; And performing an equalization operation on the pixels excluding the pixel to be preserved among the pixels.

Description

[0001] The present invention relates to a method and apparatus for image processing using a histogram,

The present invention relates to an image processing method and apparatus therefor. And more particularly, to a method and apparatus for processing an image using a histogram so as to effectively identify an object to be detected in fields such as machine vision or video surveillance.

Image equalization is one of the commonly used techniques for image enhancement and can improve the sharpness of the image by redistributing the brightness of the image using a histogram.

However, in the conventional smoothing operation, the brightness of the entire image is redistributed without discriminating the foreground and the background included in the image, thereby changing information about the detection target included in the image.

Therefore, when performing the detection operation from the image subjected to the smoothing operation, there is a problem that the information about the detection target included in the image is lost and the detection performance is deteriorated.

Korean Patent Publication No. 2009-0017871

SUMMARY OF THE INVENTION The present invention is directed to a method and apparatus for performing an equalization operation only for images other than a detection object such as defects of an inspection object included in an image, .

Another object of the present invention is to provide a method for performing a sharpness operation on a detection object such as defects of an inspection object included in an image, a person or an automobile, and an apparatus for performing the method .

The technical objects of the present invention are not limited to the above-mentioned technical problems, and other technical subjects not mentioned can be clearly understood by those skilled in the art from the following description.

According to an aspect of the present invention, there is provided a method of processing an image, the method including generating a histogram using pixels included in an image, generating a histogram, Calculating a degree of variance representing a degree of variance of the candidate pixel using the in-vivo coordinates of the candidate pixel, discriminating a target pixel of the candidate pixels using the variance, And performing an equalization operation on the pixels included in the image except for the storage target pixel.

In one embodiment, the histogram includes the intra-image coordinates for each pixel included in the image, and the step of calculating the degree of variance includes calculating the degree of variance by using the coordinates included in the histogram, And calculating a dispersion degree.

In one embodiment, the step of calculating the degree of dispersion includes: mapping the candidate pixel to a virtual region corresponding to the size of the detection object using the coordinates; and calculating the degree of variance based on the number of the mapped candidate pixels And a step of calculating

In one embodiment, the method further comprises calculating a proximity representing the proximity between the candidate pixels using the in-vivo coordinates for the candidate pixel, wherein identifying the pixel to be preserved comprises: And identifying the pixel to be preserved among the candidate pixels by further using the pixel value.

In one embodiment, the calculating the proximity may include: classifying the candidate pixels into groups of adjacent pixels adjacent to each other using the coordinates, calculating a number of candidate pixels included in each group, And calculating the proximity.

According to another aspect of the present invention, there is provided an image processing apparatus including at least one processor, a memory, and a storage device loaded with the computer program loaded by the processor and executed by the processor, An instruction for extracting a candidate pixel corresponding to a brightness of a detection target to be detected from the image in the histogram, an instruction for extracting a candidate pixel corresponding to the brightness of the detection target from the image in the histogram, An instruction to identify a pixel to be preserved among the candidate pixels using the degree of variance, an instruction to identify a pixel to be preserved among the pixels included in the image, For pixels except And may include instructions to perform equalization operations.

According to the present invention as described above, the smoothing operation is performed only for the remaining images excluding the detection target included in the image, thereby maintaining the required information and enhancing the sharpness of the image.

Further, since the information on the detection target is held, even when the smoothing operation is performed and the additional detection or recognition operation is performed, the detection target can be detected without missing.

In addition, according to the present invention as described above, there is an effect that the visibility of the detection target included in the image can be improved by performing the sharpness operation on the detection target included in the image.

1 is a flowchart illustrating an image processing method of an image processing apparatus according to an embodiment of the present invention.
2 is a diagram for explaining a histogram according to an embodiment of the present invention.
3A to 3C are diagrams for explaining a proximity of a candidate pixel according to an embodiment of the present invention.
FIG. 4 is a diagram for explaining dispersion of candidate pixels according to an exemplary embodiment of the present invention. Referring to FIG.
5 is a flowchart illustrating an image processing method of an image processing apparatus according to another embodiment of the present invention.
FIG. 6 is a diagram for explaining the dispersion of candidate pixels according to another embodiment of the present invention.
7 is a block diagram of an image processing apparatus according to an embodiment of the present invention.
8 is a hardware configuration diagram of an image processing apparatus according to an embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS The advantages and features of the present invention and the manner of achieving them will become apparent with reference to the embodiments described in detail below with reference to the accompanying drawings. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. Is provided to fully convey the scope of the invention to those skilled in the art, and the invention is only defined by the scope of the claims. Like reference numerals refer to like elements throughout the specification.

Unless defined otherwise, all terms (including technical and scientific terms) used herein may be used in a sense commonly understood by one of ordinary skill in the art to which this invention belongs. Also, commonly used predefined terms are not ideally or excessively interpreted unless explicitly defined otherwise. The terminology used herein is for the purpose of illustrating embodiments and is not intended to be limiting of the present invention. In the present specification, the singular form includes plural forms unless otherwise specified in the specification.

First, terms used in the present invention are defined.

Foreground: This is the main area that contains the object to be detected in the image. For example, in the field of machine vision, the foreground may be an area containing a defect part of the inspection object, and in the field of video remote monitoring, the foreground may be a person or an automobile emerging in the surveillance area.

Background: It is the area excluding the foreground in the image.

Contiguity: An indicator of the degree of proximity between two or more pixels in an image.

Variance: An indicator of the degree of dispersion of two or more pixels included in an image.

Distance of Pixels: The number of pixels located between the A and B pixels included in the image is the distance between A and B pixels.

Hereinafter, with reference to FIGS. 1 to 4, a description will be made of a method for performing a smoothing operation excluding a defect of an inspection object included in an image, a detection object such as a person or an automobile, and the like.

1 is a flowchart illustrating an image processing method of an image processing apparatus according to an embodiment of the present invention.

Referring to FIG. 1, the image processing apparatus receives an image including a detection target (S110).

More specifically, the image processing apparatus may receive the image transmitted from the external image providing apparatus through the network or receive the image stored in the auxiliary storage apparatus in step S110, but the present invention is not limited thereto.

In operation S120, the image processing apparatus generates a histogram representing the frequency of pixels according to the intensity of the pixels included in the input image in operation S120.

In particular, the image processing apparatus according to an embodiment of the present invention may generate a histogram including in-image coordinates of each pixel included in the input image through step S110.

The coordinates of the pixels included in the histogram and the histogram generated in step S120 will be described in more detail with reference to FIG. 1 and then with reference to FIG.

The image processing apparatus extracts candidate pixels corresponding to the brightness of the detection object in the histogram generated in step S120 (S130).

More specifically, the image processing apparatus calculates a peak point at which the frequency of the pixel having the greatest frequency is within a certain range (Range) from the brightness of the detection object in the histogram, To extract candidate pixels. In this case, the image processing apparatus can extract the candidate pixel including the pixel having the highest frequency of the pixel frequency and the pixel having the brightness similar to the frequency with the highest frequency of the pixel frequency.

In step S130, the brightness of the detection object may be set by receiving the brightness of the detection object to be identified in the image from the user. In addition, the brightness of the detection object according to another embodiment of the present invention may be set by receiving the color information of the detection object to be identified in the image from the user and calculating the brightness of the detection object based on the received color information .

The predetermined range is an error range for selecting the most appropriate brightness of the detection object from the brightness of the detection object input from the user.

The candidate pixel is a candidate pixel that can be identified as a pixel constituting the detection target in the image. In other words, the candidate pixel is a pixel constituting the foreground in the image composed of foreground and background, and step S130 is a step of extracting only foreground from the image.

The image processing apparatus calculates the proximity and variance of the candidate pixels extracted in step S130 using the coordinates of the pixels included in the image included in the histogram (S140).

More specifically, in step S140, the proximity or variance calculation result of the candidate pixel of the image processing apparatus may be either High or Low.

In step S140, a method of calculating the proximity and variance of the candidate pixels will be described in more detail with reference to FIG. 1 and with reference to FIGS. 3A to 3C and 4.

The image processing apparatus identifies pixels to be preserved among the candidate pixels by using the proximity and variance of the candidate pixels calculated in step S140 (S150).

More specifically, the image processing apparatus identifies a candidate pixel having a high degree of proximity (High) and a low degree of dispersion (Low) among the candidate pixels as a storage target pixel.

In step S150, the pixel to be preserved is a pixel constituting an object to be detected in the image. In other words, the pixel to be preserved is a pixel constituting a detection object such as a defect of the inspection object, a person or an automobile, and step S150 is a step of extracting the detection object in the foreground.

In operation S160, the image processing apparatus performs an equalization operation on the pixels included in the histogram generated in operation S120, excluding the pixels to be preserved identified through operation S150.

More specifically, the image processing apparatus performs a smoothing operation to uniformly adjust a lightness value distribution for pixels other than pixels to be preserved among pixels included in the histogram. That is, the image processing apparatus performs a smoothing operation only on a region and a background excluding the foreground detection target.

Therefore, the image processing apparatus according to the embodiment of the present invention performs smoothing only on the remaining images except for defects of the inspection object included in the image, a detection object such as a person or an automobile, It is possible to improve the sharpness of the image.

The image processing apparatus performs a sharpness operation on the pixel to be preserved identified in operation S150 (S170).

More specifically, the image processing apparatus performs a sharpening operation for adjusting the contrast of the pixels, which have been identified through step S150 and are excluded from the smoothing operation in step S160. That is, the image processing apparatus performs sharpness work on the detection target included in the image.

Accordingly, the image processing apparatus according to an embodiment of the present invention performs a sharpening operation on a detection target such as a defect of the inspection object included in the image, a person, an automobile or the like, thereby improving the visibility of the detection target included in the image The effect can be demonstrated.

In operation S180, the image processing apparatus reconstructs the image including the pixel subjected to the smoothing operation in step S160 and the preservation target pixel on which the sharpening operation has been performed in step S170.

Hereinafter, the coordinates of the pixels included in the histogram and the histogram generated in step S120 will be described in detail with reference to FIG.

2 is a diagram for explaining a histogram according to an embodiment of the present invention.

Referring to FIG. 2, a histogram according to an exemplary embodiment of the present invention has a brightness of a pixel as an X-axis and a frequency of pixels as a Y-axis. The histogram has a stacked bar graph shape capable of expressing the frequency of pixels having a specific brightness among the pixels included in the image.

In particular, as shown in FIG. 2, when each pixel included in an image is represented by one block (Block, 25) constituting a histogram stacking bar graph, each block 25 in the histogram is represented by a corresponding pixel An identifier 27 for identifying the pixel, and an x, y coordinate 29 in the image of the pixel.

Then, the image processing apparatus calculates the brightness 21 having the largest frequency of the pixels among the brightness within the certain range 23 from the brightness of the detection object in the histogram, and calculates the brightness 21 having the greatest frequency of the pixels, A candidate pixel can be extracted including a pixel having a brightness similar to the brightness 21 having the highest frequency of the pixel.

Therefore, the image processing apparatus according to the embodiment of the present invention includes the coordinates of the pixels in the histogram, and performs operations necessary for image processing using only the coordinates of the pixels included in the histogram without referring to the input image The effect can be achieved.

Hereinafter, the method of calculating the proximity and variance of the candidate pixels in step S140 will be described in detail with reference to FIGS. 3A to 3C and FIG.

3A to 3C are diagrams for explaining a proximity of a candidate pixel according to an embodiment of the present invention.

Referring to FIG. 3A, each candidate pixel includes an identifier for identifying a pixel and x and y coordinates of the pixel in the image.

In this case, the distance between x0, y0 of the C0 candidate pixel and x1, y1 of the C1 candidate pixel is 0, and x4, y4 of the C4 candidate pixel, x5, y5 of the C4 candidate pixel, y7 of the candidate pixel, Assume that the distance between x8 and y8 of the C8 candidate pixel is zero.

Referring to FIG. 3B, candidate pixels may be distinguished into one or more groups of candidate pixels having a distance of zero.

Therefore, the image processing apparatus classifies the candidate pixels into groups of adjacent candidate pixels, respectively.

That is, the image processing apparatus includes a group 1 (31) composed of C0 candidate pixels and C1 candidate pixels, a group 2 (33) composed of C2 candidate pixels, a group 3 (35) composed of C3 candidate pixels, , A group 37 composed of a C4 candidate pixel, a group C5 candidate pixel, a group C7 candidate pixel, and a group C8 candidate pixel, and a group 4 39 composed of a C6 candidate pixel.

Referring to FIG. 3C, the candidate pixels may be represented by group 1 (31) to group 4 (39) including each candidate pixel.

The image processing apparatus calculates the number of candidate pixels included in the group 1 (31) to the group 4 (39).

If the number of candidate pixels included in each group is larger than the minimum size of the detection target, the image processing apparatus calculates the proximity of the candidate pixels included in the group to be high. When the number of candidate pixels included in each group is smaller than the minimum size of the detection target, the image processing apparatus calculates the proximity of the candidate pixels included in the group as low.

Here, the minimum size of the detection target may be set by receiving a minimum size of the detection target to be identified in the image from the user, but it is not limited thereto and may be preset in the image processing apparatus.

FIG. 4 is a diagram for explaining dispersion of candidate pixels according to an exemplary embodiment of the present invention. Referring to FIG.

Referring to FIG. 4, the image processing apparatus maps candidate pixels 43 and 45 to a two-dimensional kernel (Kernel) having the minimum size 41 of the detection object as one side based on the in-image coordinates of the candidate pixel . Here, the kernel is a kind of virtual area for convenience of calculation.

When the candidate pixel mapped to each of the kernels satisfies Equation (1), the image processing apparatus calculates the degree of variance of the candidate pixels included in the corresponding kernel as High and does not satisfy Equation (1) , The degree of dispersion of the candidate pixels included in the kernel is calculated as Low.

(Where N is the minimum size of the object to be detected and P is the number of candidate pixels mapped to the kernel)

However, when the candidate pixels mapped to one kernel are classified into two groups according to the coordinates in the image, the image processing apparatus calculates the number P of candidate pixels mapped to the corresponding kernel as follows.

For convenience of explanation, let us say that two groups mapped to one kernel are group A (43) and group B (45), respectively.

The candidate pixel A 47 closest to the group B 45 among the candidate pixels included in the group A 43 and the candidate pixel B 49 49 closest to the group A 43 among the candidate pixels included in the group B 45 ) Is greater than N , the image processing apparatus calculates P as the larger of the number of candidate pixels included in the group A and the number of candidate pixels included in the group B. [

Conversely, among the candidate pixels included in the group A 43, the candidate pixel A 47 closest to the group B 45 and the candidate pixel 46 closest to the group A 43 among the candidate pixels included in the group B 45 If the distance of B (49) is less than or equal to N , the image processing apparatus calculates P as the sum of the number of candidate pixels included in group A and the number of candidate pixels included in group B.

Hereinafter, with reference to FIG. 5 and FIG. 6, a method for performing a smoothing operation including a detection object such as fog or smoke included in an image will be described.

5 is a flowchart illustrating an image processing method of an image processing apparatus according to another embodiment of the present invention.

Referring to FIG. 5, the image processing apparatus receives an image including a detection target (S210).

More specifically, the image processing apparatus receives the image transmitted from the external image providing apparatus through the network or receives the image stored in the auxiliary storage apparatus in step S210, but is not limited thereto.

In operation S220, the image processing apparatus generates a histogram representing the frequency of pixels according to the intensity of the pixels included in the input image in operation S210.

In particular, the image processing apparatus according to an embodiment of the present invention can generate a histogram by including in-image coordinates of each pixel included in the input image through the histogram in step S210.

The detailed description of the coordinates of the pixels included in the histogram and the histogram generated in step S220 is as described above with reference to FIG.

The image processing apparatus extracts a candidate pixel corresponding to the brightness of the detection object in the histogram generated in step S220 (S230).

In step S230, the candidate pixel is a candidate pixel that can be identified as a pixel constituting the detection target in the image. In other words, the candidate pixel is a pixel constituting the foreground in the image composed of foreground and background, and step S130 is a step of extracting only foreground from the image.

The image processing apparatus calculates the variance of the candidate pixels extracted in step S230 using the coordinates of the pixels included in the image included in the histogram (S240).

More specifically, when the detection object contained in the image is in the form of fog, smoke, or the like, the proximity of the candidate pixel is not separately calculated, and only the degree of dispersion of the candidate pixel is calculated. The method of calculating the variance of the candidate pixels in step S240 will be described in more detail with reference to FIG. 5 and with reference to FIG.

The image processing apparatus identifies a pixel to be preserved among the candidate pixels using the degree of variance of the candidate pixel calculated in step S240 (S250).

More specifically, the image processing apparatus identifies a candidate pixel having a high degree of dispersion (Candidate) among candidate pixels as a target pixel to be stored. Then, the image processing apparatus identifies candidate pixels having low dispersion (Low) among the candidate pixels as pixels constituting the detection target.

In step S250, the pixels constituting the detection target are pixels constituting an object to be detected such as fog or smoke.

The pixel to be preserved is a pixel excluding the pixels constituting the detection target among the pixels included in the foreground. In other words, the pixel to be preserved may be a defect constituting a defect of the inspection object, a pixel constituting a person or an automobile, etc., in which fog or smoke is excluded from the foreground.

In operation S260, the image processing apparatus performs a smoothing operation on the pixels included in the histogram generated in operation S220, excluding the pixels to be preserved identified through operation S250.

More specifically, the image processing apparatus performs a smoothing operation to uniformly adjust a lightness value distribution for pixels other than pixels to be preserved among pixels included in the histogram. That is, the image processing apparatus performs smoothing only on detection objects and backgrounds such as fog or smoke included in the foreground.

Therefore, the image processing apparatus according to an embodiment of the present invention performs smoothing only on the detection object and the background of the image such as fog or smoke included in the image, thereby maintaining necessary information and improving the sharpness of the image The effect can be demonstrated.

In operation S270, the image processing apparatus reconstructs the image including the pixels subjected to the smoothing operation and the pixels not subjected to the smoothing operation in operation S260.

Hereinafter, a method of calculating the variance of candidate pixels in step S240 will be described in detail with reference to FIG.

FIG. 6 is a diagram for explaining the dispersion of candidate pixels according to another embodiment of the present invention.

Referring to FIG. 6, the image processing apparatus maps a candidate pixel 63 to a two-dimensional kernel (Kernel) having the minimum size 61 of the detection object as one side based on the in-image coordinates of the candidate pixel. Here, the kernel is a kind of virtual area for convenience of calculation.

When the candidate pixel mapped to each of the kernels satisfies Equation (2), the image processing apparatus calculates the degree of variance of the candidate pixels included in the corresponding kernel as High and does not satisfy Equation (2) , The degree of dispersion of the candidate pixels included in the kernel is calculated as Low.

Hereinafter, the logical configuration of the image processing apparatus according to an embodiment of the present invention will be described in detail with reference to FIGS. 7 and 8. FIG.

7 is a block diagram of an image processing apparatus 300 according to an embodiment of the present invention.

7, the image processing apparatus 300 includes a communication unit 310, an input / output unit 320, a histogram generation unit 330, a candidate pixel identification unit 340, a proximity calculation unit 350, A storage unit 360, a storage object identification unit 370, an image processing unit 380, and an image reconstruction unit 390. [

The communication unit 310 transmits / receives data to / from an external image providing apparatus through a network, receives an original image for image processing, or transmits a processed image.

The input / output unit 320 inputs and outputs data necessary for the image processing apparatus from the user, receives the brightness or color information of the detection target, and the minimum size of the detection target.

The histogram generator 330 generates a histogram representing the frequency of pixels according to brightness of the pixel using the image received through the communication unit 310. [ In particular, the histogram generator 330 according to an exemplary embodiment of the present invention may generate a histogram including the in-image coordinates of each pixel included in the image.

The candidate pixel identification unit 340 extracts a candidate pixel corresponding to the brightness of the detection subject input through the input / output unit 320 in the histogram generated through the histogram generation unit 330. [

More specifically, the candidate pixel identification unit 340 calculates a brightness with the greatest frequency of pixels among the brightness within a certain range from the brightness of the detection object in the histogram, and includes pixels having the greatest frequency of the pixels Extract candidate pixels. In this case, the candidate pixel identification unit 340 may extract the candidate pixel including the pixel having the highest frequency of the pixel frequency and the pixel having the brightness similar to the frequency with the highest frequency of the pixel frequency.

The proximity calculator 350 calculates the proximity of the candidate pixel extracted through the candidate pixel identifier 340. More specifically, the proximity calculation method for the candidate pixel of the proximity calculation unit 350 is as described above with reference to FIGS. 3A to 3C.

The variance calculation unit 360 calculates the variance of the candidate pixels extracted through the candidate pixel identification unit 340. More specifically, the method for calculating the variance of the candidate pixels of the variance calculation unit 360 is as described above with reference to FIGS. 4 and 6. FIG.

The storage target identification unit 370 identifies the storage target pixel among the candidate pixels. More specifically, when a smoothing operation is to be performed except for a detection target, the preservation object identification unit 370 identifies candidate pixels having a high degree of proximity (High) and a low degree of dispersion (Low) .

Alternatively, when a smoothing operation is to be performed including a detection target, the preservation object identification unit 370 identifies a candidate pixel having a high degree of dispersion among the candidate pixels as a preservation object pixel.

The image processing unit 380 performs smoothing on the pixels other than the preservation target pixels identified through the preservation object identification unit 370 among the pixels included in the histogram generated through the histogram generation unit 330. [ In addition, the image processing unit 380 may further perform a sharpness operation on the storage target pixel identified through the storage target identification unit 370. [

The image reconstructing unit 390 reconstructs the image including the pixels subjected to the smoothing operation and the pixels not subjected to the smoothing operation through the image processing unit 380.

Each of the components in FIG. 7 may refer to software or hardware such as an FPGA (Field-Programmable Gate Array) or an ASIC (Application-Specific Integrated Circuit). However, the components are not limited to software or hardware, and may be configured to be addressable storage media, and configured to execute one or more processors. The functions provided in the components may be implemented by a more detailed component or may be implemented by a single component that performs a specific function by combining a plurality of components.

8 is a hardware block diagram of the image processing apparatus 300 according to an embodiment of the present invention.

The image processing apparatus 300 includes a processor 410 for executing a command, a RAM 420, a storage 430 for storing computer program data in which an image processing method is implemented, a network interface 440 for transmitting / An I / O 450 and a processor 410, a RAM 420, a storage 430, a network interface 440, and an I / O 450 (not shown) for inputting and receiving data necessary for operation of the image processing apparatus from a user And a data bus 460 connected to the data bus 460 as a data movement path.

The storage 430 may store data such as an executable file and a resource file for execution of the computer program. More specifically, the storage 430 includes a series of instructions for generating a histogram using pixels included in the image, instructions for extracting candidate pixels corresponding to the brightness of the detection object to be detected from the image in the histogram, An instruction to calculate a degree of variance indicating a degree of variance of the candidate pixel using the in-vivo coordinate of the candidate pixel; an instruction to identify a candidate pixel to be preserved among the candidate pixels using the variance; And performing an equalization operation on pixels other than the pixel to be preserved among the pixels to be preserved.

While the present invention has been described in connection with what is presently considered to be practical exemplary embodiments, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, You will understand. It is therefore to be understood that the above-described embodiments are illustrative in all aspects and not restrictive.

Claims

Generating a histogram using pixels included in the image;
Extracting a candidate pixel corresponding to a brightness of a detection target to be detected from the image in the histogram;
Calculating a degree of variance representing a degree of variance of the candidate pixel using the intra-image coordinates of the candidate pixel;
Identifying a pixel to be preserved among the candidate pixels using the degree of dispersion; And
And performing an equalization operation on pixels other than the pixel to be preserved among the pixels included in the image.

The method according to claim 1,
The histogram may include:
The in-image coordinates for each pixel included in the image,
Wherein the step of calculating the degree of dispersion comprises:
And calculating a variance of the candidate pixel using the coordinates included in the histogram.

The method according to claim 1,
Wherein the step of calculating the degree of dispersion comprises:
Mapping the candidate pixel to a virtual area corresponding to the size of the detection object using the coordinates, and calculating the variance based on the number of the mapped candidate pixels.

The method according to claim 1,
Further comprising the step of calculating a degree of proximity indicating the degree of proximity between the candidate pixels using the in-vivo coordinates for the candidate pixel,
Wherein the identifying of the pixel to be preserved comprises:
And identifying the pixel to be preserved among the candidate pixels by further using the proximity.

5. The method of claim 4,
The step of calculating the proximity may comprise:
And calculating the proximity by classifying the candidate pixels into groups of adjacent candidate pixels using the coordinates and comparing the number of candidate pixels included in each group and the size of the detection object. Image processing method.

One or more processors;
Memory; And
A storage device loaded with the computer program recorded by the processor and executed by the processor,
The computer program comprising:
A series of instructions for generating a histogram using pixels contained in an image;
An instruction for extracting a candidate pixel corresponding to a brightness of a detection target to be detected from the image in the histogram;
An instruction to calculate a degree of variance indicating a degree of variance of the candidate pixel using the in-vivo coordinates of the candidate pixel;
An instruction for identifying a pixel to be preserved among the candidate pixels using the degree of dispersion; And
And an instruction for performing an equalization operation on pixels other than the storage target pixel among the pixels included in the image.