A solar cell shading rate quick and efficient testing method, system and device

Abstract

This invention discloses a rapid and efficient method for testing the shading rate of solar cells, comprising: image processing of the target grayscale image; converting the reference grayscale image into a reference gradient histogram; converting the processed target grayscale image into a target gradient histogram; comparing the reference gradient histogram and the target gradient histogram to extract target features, wherein the target features include target edges and target internal feature units; counting the number of pixels in the region of the target edge to obtain the total number of pixels; counting the number of pixels in the target internal feature units to obtain the number of printed pattern pixels; and calculating the shading rate based on the total number of pixels and the number of printed pattern pixels. Accordingly, this invention also discloses a testing system and computer equipment related to the above testing method. Using this invention, testing efficiency is high, the repeatability and accuracy of test results are improved, and testing manpower and time costs are reduced.

Description

Technical Field

[0001] This invention relates to the field of solar power generation technology, and in particular to a fast and efficient method, system and equipment for testing the shading rate of solar cells. Background Technology

[0002] One existing method for testing the shading efficiency of solar cells is titled "Test Method for Shading Efficiency of Solar Cell Grid Lines," application number CN201811602273.8. This technical solution involves: using a quantum efficiency meter to test the EQE value. The quantum efficiency meter emits a light spot of a specific wavelength, and the position of the light spot is adjusted to be parallel to the grid line to obtain the EQE value of that light spot; the wavelength of the light spot is adjusted within a specific wavelength range to obtain a series of first EQE values; the quantum efficiency meter emits a light spot of a specific wavelength, and the position of the light spot is adjusted to be perpendicular to the grid line, and the wavelength of the light spot is adjusted within a specific wavelength range to obtain a series of second EQE values; the first EQE values ​​are subtracted from the second EQE values ​​to obtain a series of differences; the average of these differences is calculated to obtain the shading efficiency.

[0003] This type of technical solution mainly involves offline testing. Measuring technicians use 2D measuring instruments or optical equipment such as Zeta microscopes to perform multi-piece, multi-point tests on the printed finished batteries to obtain the printed linewidth and roughly estimate the shading rate based on the graphic design. This process requires adjusting the light spot direction twice and testing data at multiple points, making it complex, time-consuming, and highly susceptible to the effects of acquisition gradient and line uniformity, resulting in poor accuracy and repeatability of the results. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a fast and efficient test method for the shading rate of solar cells, which has high testing efficiency, improves the repeatability and accuracy of test results, and reduces testing manpower and time costs.

[0005] To address the aforementioned technical problems, this invention provides a rapid and efficient method for testing the shading rate of solar cells, comprising: acquiring a reference grayscale image; acquiring a target grayscale image of the solar cell; performing image processing on the target grayscale image; converting the reference grayscale image into a reference gradient histogram; converting the processed target grayscale image into a target gradient histogram; comparing the reference gradient histogram and the target gradient histogram to extract target features, wherein the target features include target edges and target internal feature units; counting the number of pixels in the target edge region to obtain the total number of pixels; counting the number of pixels in the target internal feature units to obtain the number of printed pattern pixels; and calculating the shading rate based on the total number of pixels and the number of printed pattern pixels.

[0006] As an improvement to the above scheme, the step of image processing the target grayscale image includes: enhancing and sharpening the target grayscale image; and filtering and denoising the enhanced and sharpened grayscale image.

[0007] As an improvement to the above scheme, the step of converting the reference grayscale image into a reference gradient histogram includes: calculating the gradient of the reference grayscale image; segmenting the reference grayscale image to form a number of image units; and calculating the gradient histogram of each image unit according to the magnitude and direction of the gradient to obtain the reference gradient histogram.

[0008] As an improvement to the above scheme, the step of comparing the benchmark gradient histogram and the target gradient histogram to extract target features includes: comparing the benchmark gradient histogram and the target gradient histogram and removing overlapping gradient histograms to obtain a target feature gradient histogram; calculating the target features based on the target feature gradient histogram; obtaining all RGB values ​​of the target features, and distinguishing the target edges and each target internal feature unit in the target grayscale image based on the corresponding RGB values.

[0009] As an improvement to the above scheme, the target grayscale image includes a front target grayscale image and a back target grayscale image, and the occlusion rate includes a front occlusion rate and a back occlusion rate.

[0010] As an improvement to the above solution, the step of calculating the occlusion rate based on the total number of pixels and the number of pixels in the printed pattern includes: calculating the occlusion rate according to formula F. shading = (a+b+c) / A, calculate the front shading rate; according to the formula R shading = (d+e+f+g) / A, calculate the back shading rate; where F shading Let be the front occlusion rate, 'a' be the number of pixels in the front target border, 'b' be the number of pixels in the front target main grid, 'c' be the number of pixels in the front target fine grid, 'A' be the total number of pixels, and 'R' be the total number of pixels. shading denoted as the back-side occlusion rate, d as the number of pixels in the back-side target border, e as the number of pixels in the back-side target main gate, f as the number of pixels in the back-side target fine gate, and g as the number of pixels in the back-side target electrode.

[0011] Accordingly, the present invention also provides a fast and efficient testing system for the shading rate of solar cells, comprising: a reference acquisition module for acquiring a reference grayscale image; a target acquisition module for acquiring a target grayscale image of the solar cell; an image processing module for performing image processing on the target grayscale image; an orientation gradient histogram conversion module for converting the reference grayscale image into a reference gradient histogram and converting the processed target grayscale image into a target gradient histogram; an overlap comparison module for comparing the reference gradient histogram and the target gradient histogram to extract target features, wherein the target features include target edges and target internal feature units; a pixel counting module for counting the number of pixels in the region of the target edge to obtain the total number of pixels and counting the number of pixels in the target internal feature units to obtain the number of printed pattern pixels; and a calculation module for calculating the shading rate based on the total number of pixels and the number of printed pattern pixels.

[0012] As an improvement to the above solution, the image processing module includes an enhancement and sharpening unit and a filtering and noise reduction unit; the enhancement and sharpening unit is used to enhance and sharpen the target grayscale image; the filtering and noise reduction unit is used to filter and reduce noise in the enhanced and sharpened grayscale image.

[0013] As an improvement to the above scheme, the directional gradient histogram conversion module includes a gradient calculation unit, a segmentation unit, and a gradient histogram construction unit; the gradient calculation unit is used to calculate the gradient of the reference grayscale image or the target grayscale histogram; the segmentation unit is used to segment the reference grayscale image or the target grayscale histogram to form several image units; the gradient histogram construction unit is used to calculate the gradient histogram of each image unit according to the magnitude and direction of the gradient to obtain the reference gradient histogram or the target gradient histogram.

[0014] As an improvement to the above scheme, the overlap comparison module includes a comparison unit, a feature calculation unit, and a feature extraction unit; the comparison unit is used to compare the benchmark gradient histogram and the target gradient histogram and remove the overlapping gradient histograms to obtain the target feature gradient histogram; the feature calculation unit is used to calculate the target feature based on the target feature gradient histogram; the feature extraction unit is used to obtain all RGB values ​​of the target feature and distinguish the target edge and each target internal feature unit in the target grayscale image based on the corresponding RGB values.

[0015] Accordingly, the present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, characterized in that the processor executes the computer program to implement the steps of the above-described testing method.

[0016] Compared with the prior art, the beneficial effects of implementing the present invention are as follows:

[0017] This invention provides a fast and efficient method for testing the shading rate of solar cells. Since the number of pixels in the printed pattern is the sum of the number of pixels in all the internal feature units of the target, the number of pixels in the printed pattern directly reflects all the paste in the printed pattern, including the paste on the grid lines and the scattered paste near the grid lines. Therefore, the shading rate obtained by the test is more accurate, improving the repeatability and accuracy of the test results. Moreover, it eliminates the need to set up multiple sets of control tests and point tests, resulting in high testing efficiency and reducing testing manpower and time costs.

[0018] Secondly, real-time monitoring through appropriate testing systems or computer equipment provides data support for establishing a theoretical model between printing efficiency and light-blocking rate. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the implementation of the rapid and efficient testing method for the shading rate of solar cells according to the present invention.

[0020] Figure 2 This is a schematic diagram of the structure of the front reference grayscale image in this invention;

[0021] Figure 3 This is a schematic diagram of the structure of the back reference grayscale image in this invention;

[0022] Figure 4 yes Figure 1 Flowchart of image processing for a target grayscale image;

[0023] Figure 5 yes Figure 1 The flowchart illustrates the implementation process of comparing a frontal digital image with a sample frontal image to extract the edge and frontal target features of the target silicon wafer;

[0024] Figure 6 yes Figure 1 The flowchart illustrates the implementation process of converting a baseline grayscale image into a baseline gradient histogram.

[0025] Figure 7 yes Figure 1 The flowchart shows the implementation process of converting the processed target grayscale image into a target gradient histogram.

[0026] Figure 8 This is a schematic diagram of the principle of the solar cell shading rate testing system of the present invention;

[0027] Figure 9 yes Figure 8 Block diagram of the image processing module;

[0028] Figure 10 yes Figure 8 A block diagram illustrating the principle of the directional gradient histogram conversion module;

[0029] Figure 11 yes Figure 8 Block diagram of the overlap comparison module. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings.

[0031] like Figure 1 As shown, Figure 1 The real-time workflow of a rapid and efficient method for testing the shading rate of solar cells is shown, including:

[0032] S101, acquire the baseline grayscale image.

[0033] The reference grayscale image is a pre-drawn sample image of a solar cell printing, stored in a corresponding memory. The reference grayscale image includes a front reference grayscale image and a back reference grayscale image. For example... Figure 2 As shown, the front reference grayscale image includes a reference edge 1, a front reference border 2, a front reference main grid 3, and a front reference fine grid 4; as Figure 3 As shown, the back reference grayscale image includes a reference silicon wafer edge 1, a back reference frame 5, a back reference main gate 6, a back reference fine gate 7, and a back reference electrode.

[0034] S102, acquire the target grayscale image of the solar cell.

[0035] The target grayscale image of the solar cell is sampled using a CCD camera through optical scanning. The target grayscale image includes a front target grayscale image and a back target grayscale image.

[0036] S103, perform image processing on the target grayscale image.

[0037] S104, convert the reference grayscale image into a reference gradient histogram.

[0038] The reference gradient histogram includes a front reference gradient histogram and a back reference gradient histogram.

[0039] S105 converts the processed target grayscale image into a target gradient histogram.

[0040] The target gradient histogram includes a front target gradient histogram and a back target gradient histogram.

[0041] S106, the benchmark gradient histogram and the target gradient histogram are overlapped and compared to extract target features, the target features including target edges and target internal feature units.

[0042] The target internal feature units include front target internal feature units and rear target internal feature units.

[0043] When calculating the front shading rate, the front reference gradient histogram and the front target gradient histogram are overlapped and compared to extract the target edge and the front target internal feature units. The front target internal feature units include the front target border, the front target main grid, and the front target fine grid.

[0044] When calculating the back shading rate, the back reference gradient histogram and the back target gradient histogram are overlapped and compared to extract the target edge and back target features. The back target internal feature unit includes the back target border, the back target main grid, the back target fine grid, and the back target electrode.

[0045] S107, count the number of pixels in the region of the target edge to obtain the total number of pixels.

[0046] The target edge refers to the edge of the raw silicon wafer, therefore the total number of pixels refers to the total number of pixels on the raw silicon wafer.

[0047] S108, count the number of pixels in the feature unit inside the target to obtain the number of pixels in the printed pattern.

[0048] When calculating the front shading rate, the number of pixels in the printed pattern is the sum of the number of pixels in the feature units inside the front target.

[0049] When calculating the back-side occlusion rate, the number of pixels in the printed pattern is the sum of the number of pixels in the internal feature units of the back-side target.

[0050] S109, calculate the shading rate based on the total number of pixels and the number of pixels in the printed pattern.

[0051] When calculating the front shading rate, according to the formula

[0052] F shading = (a+b+c) / A

[0053] Calculate the front shading rate;

[0054] Among them, F shading denoted as front occlusion rate, where a is the number of pixels on the front target border, b is the number of pixels on the front target main grid, c is the number of pixels on the front target fine grid, and A is the total number of pixels.

[0055] When calculating the back shading rate, according to the formula

[0056] R shading =(d+e+f+g) / A

[0057] Calculate the back shading rate;

[0058] Among them, R shading denoted as the back-side occlusion rate, d as the number of pixels on the back-side target border, e as the number of pixels on the back-side target main gate, f as the number of pixels on the back-side target fine gate, g as the number of pixels on the back-side target electrode, and A as the total number of pixels.

[0059] It should be noted that the present invention provides a fast and efficient method for testing the shading rate of solar cells. Since the number of pixels in the printed pattern is the sum of the number of pixels in all the internal feature units of the target, the number of pixels in the printed pattern directly reflects all the paste in the printed pattern, including the paste on the grid lines and the scattered paste near the grid lines. Therefore, the shading rate obtained by the test is more accurate, improving the repeatability and accuracy of the test results. Moreover, it eliminates the need to set up multiple sets of control tests and point tests, resulting in high testing efficiency and reducing testing manpower and time costs. Secondly, real-time monitoring through a corresponding testing system or computer equipment provides data support for establishing a theoretical model between printing efficiency and shading rate.

[0060] Specifically, such as Figure 4 As shown, step S103 includes:

[0061] S301, the target grayscale image is enhanced and sharpened.

[0062] The binary method is used to perform image enhancement and sharpening processes on the acquired digital images to improve the comparison accuracy of the detected features.

[0063] S302 performs filtering and noise reduction processing on the enhanced and sharpened grayscale image.

[0064] Median filtering is used to reduce noise in the binarized digitized image in order to enhance image recognition.

[0065] Specifically, such as Figure 5 As shown, step S104 includes:

[0066] S401, Calculate the gradient of the reference grayscale image.

[0067] The Sobel operator with a kernel size of 1 is used to compute the x- and y-gradients of the baseline grayscale image, and then the combined gradient of the x- and y-gradients, including magnitude and value, is calculated. At each pixel, the gradient has both magnitude and direction. The x-gradient enhances vertical edge features, and the y-gradient enhances horizontal edge features. This ensures that useful features (contours) are preserved while irrelevant and unimportant information is removed.

[0068] S402, the reference grayscale image is segmented to form several image units.

[0069] Because the effective features in a whole gradient image are very sparse, the computation is not only large, but the effect may also be poor. Therefore, it is necessary to divide the reference grayscale image into multiple image units based on the grayscale threshold segmentation method of metric space.

[0070] S403, calculate the gradient histogram of each image unit based on the magnitude and direction of the gradient to obtain the reference gradient histogram.

[0071] Each pixel in each image unit has a corresponding gradient direction (angle) and magnitude (amplitude). Adding the gradient values ​​of all pixels to the bin corresponding to their respective angles forms the histogram of that image unit. Once the histograms of all image units are constructed, the baseline gradient histogram is obtained.

[0072] Specifically, such as Figure 6 As shown, step S105 includes:

[0073] S501, Calculate the gradient of the target grayscale image.

[0074] The principle is the same as step S401 above.

[0075] S502, the target grayscale image is segmented to form several image units.

[0076] The principle is the same as step S402 above.

[0077] S503, calculate the gradient histogram of each image unit based on the magnitude and direction of the gradient to obtain the target gradient histogram.

[0078] The principle is the same as step S403 above.

[0079] Specifically, such as Figure 7 As shown, step S106 includes:

[0080] S601, compare the reference gradient histogram and the target gradient histogram and remove the overlapping gradient histograms to obtain the target feature gradient histogram.

[0081] S602, the target features are calculated based on the target feature gradient histogram.

[0082] S603, obtain all RGB values ​​of the target feature, and distinguish the target edge and each target internal feature unit in the target grayscale image according to the corresponding RGB values.

[0083] Different feature units have different RGB values. For example, the RGB value of the front target main grid is A. The front target main grid can be extracted by obtaining all pixels with an RGB value of A in the target image, which also facilitates the pixel statistics in steps S107 and S108.

[0084] like Figure 8 As shown, the present invention also provides a fast and efficient testing system for the shading rate of solar cells, comprising:

[0085] The reference acquisition module 10 is used to input a reference grayscale image;

[0086] Target acquisition module 20 is used to acquire a grayscale image of the target solar cell;

[0087] Image processing module 30 is used to perform image processing on the target grayscale image;

[0088] The directional gradient histogram conversion module 40 is used to convert the reference grayscale image into a reference gradient histogram and to convert the processed target grayscale image into a target gradient histogram.

[0089] The overlap comparison module 50 is used to compare the benchmark gradient histogram and the target gradient histogram to extract target features, the target features including target edges and target internal feature units;

[0090] The pixel counting module 60 is used to count the number of pixels in the region of the target edge to obtain the total number of pixels and to count the number of pixels in the feature units inside the target to obtain the number of pixels in the printed pattern.

[0091] The calculation module 70 is used to calculate the occlusion rate based on the total number of pixels and the number of pixels in the printed pattern.

[0092] The reference acquisition module 10 is connected to the overlap comparison module 50 through the orientation gradient histogram conversion module 40, the target acquisition module 20 is connected to the overlap comparison module 50 through the image processing module 30 and the orientation gradient histogram conversion module 40 in sequence, and the overlap comparison module 50 is connected to the calculation module 70 through the pixel point statistics module 60.

[0093] like Figure 9As shown, the image processing module 30 includes an enhancement and sharpening unit 301 and a filtering and noise reduction unit 302 connected to each other; the enhancement and sharpening unit 301 is used to enhance and sharpen the target grayscale image; the filtering and noise reduction unit 302 is used to filter and reduce noise in the enhanced and sharpened grayscale image.

[0094] like Figure 10 As shown, the directional gradient histogram conversion module 40 includes a gradient calculation unit 401, a segmentation unit 402, and a gradient histogram construction unit 403 connected in sequence. The gradient calculation unit 401 is used to calculate the gradient of the reference grayscale image or the target grayscale histogram. The segmentation unit 402 is used to segment the reference grayscale image or the target grayscale histogram to form a number of image units. The gradient histogram construction unit 403 is used to calculate the gradient histogram of each image unit according to the magnitude and direction of the gradient to obtain the reference gradient histogram or the target gradient histogram.

[0095] like Figure 11 As shown, the overlap comparison module 50 includes a comparison unit 501, a feature calculation unit 502, and a feature extraction unit 503 connected in sequence. The comparison unit 501 is used to compare the reference gradient histogram and the target gradient histogram and remove the overlapping gradient histograms to obtain the target feature gradient histogram. The feature calculation unit 502 is used to calculate the target feature based on the target feature gradient histogram. The feature extraction unit 503 is used to obtain all RGB values ​​of the target feature and distinguish the target edge and each target internal feature unit in the target grayscale image based on the corresponding RGB values.

[0096] Accordingly, the present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, characterized in that the processor executes the computer program to implement the steps of the above-described testing method.

[0097] Therefore, by adopting this invention, the testing efficiency is high, the repeatability and accuracy of the test results are improved, the testing manpower and time costs are reduced, and real-time monitoring through corresponding testing systems or computer equipment provides data support for establishing a theoretical model between printing efficiency and light blocking rate.

[0098] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A rapid and efficient method for testing the shading rate of solar cells, characterized in that, include: Obtain the baseline grayscale image; Acquire the target grayscale image of the solar cell; Perform image processing on the target grayscale image; The reference grayscale image is converted into a reference gradient histogram; Convert the processed target grayscale image into a target gradient histogram; The benchmark gradient histogram and the target gradient histogram are overlapped and compared to extract target features. The target features include target edges and target internal feature units. The target edges are the edges of the raw silicon wafer. The target internal feature units include front target internal feature units and back target internal feature units. The front target internal feature units include a front target border, a front target main gate, and a front target fine gate. The back target internal feature units include a back target border, a back target main gate, a back target fine gate, and a back target electrode. The total number of pixels is obtained by counting the number of pixels within the region of the target edge. The number of pixels in the feature units inside the target is counted to obtain the number of pixels in the printed pattern; The occlusion rate is calculated based on the total number of pixels and the number of pixels in the printed graphic. The target grayscale image includes a front target grayscale image and a back target grayscale image, and the occlusion rate includes a front occlusion rate and a back occlusion rate. The step of calculating the occlusion rate based on the total number of pixels and the number of pixels in the printed graphic includes: According to formula F shading = (a+b+c) / A, calculate the front shading rate; According to formula R shading = (d+e+f+g) / A, calculate the back shading rate; Among them, F shading Let be the front occlusion rate, 'a' be the number of pixels in the front target border, 'b' be the number of pixels in the front target main grid, 'c' be the number of pixels in the front target fine grid, 'A' be the total number of pixels, and 'R' be the total number of pixels. shading denoted as the back-side occlusion rate, d as the number of pixels in the back-side target border, e as the number of pixels in the back-side target main gate, f as the number of pixels in the back-side target fine gate, and g as the number of pixels in the back-side target electrode.

2. The rapid and efficient test method for solar cell shading rate according to claim 1, characterized in that, The steps for image processing of the target grayscale image include: The target grayscale image is enhanced and sharpened. The enhanced and sharpened grayscale image is then filtered and denoised.

3. The rapid and efficient test method for solar cell shading rate according to claim 1, characterized in that, The step of converting the reference grayscale image into a reference gradient histogram includes: Calculate the gradient of the reference grayscale image; The reference grayscale image is segmented to form several image units; The gradient histogram of each image unit is calculated based on the magnitude and direction of the gradient to obtain the baseline gradient histogram.

4. The rapid and efficient test method for solar cell shading rate according to claim 3, characterized in that, The step of comparing and aligning the baseline gradient histogram and the target gradient histogram to extract target features includes: The baseline gradient histogram and the target gradient histogram are compared and the overlapping gradient histograms are removed to obtain the target feature gradient histogram. The target features are calculated based on the target feature gradient histogram. All RGB values ​​of the target feature are obtained, and the target edge and each target internal feature unit are distinguished in the target grayscale image according to the corresponding RGB values.

5. A rapid and efficient testing system for the shading rate of solar cells, characterized in that, include: The benchmark acquisition module is used to acquire a benchmark grayscale image; The target acquisition module is used to acquire the target grayscale image of the solar cell; The image processing module is used to perform image processing on the target grayscale image; The directional gradient histogram conversion module is used to convert the reference grayscale image into a reference gradient histogram and to convert the processed target grayscale image into a target gradient histogram. The overlap comparison module is used to overlap and compare the reference gradient histogram and the target gradient histogram to extract target features. The target features include target edges and target internal feature units. The target edges are the edges of the raw silicon wafer. The target internal feature units include front target internal feature units and back target internal feature units. The front target internal feature units include a front target border, a front target main gate, and a front target fine gate. The back target internal feature units include a back target border, a back target main gate, a back target fine gate, and a back target electrode. The pixel counting module is used to count the number of pixels in the region of the target edge to obtain the total number of pixels and to count the number of pixels in the feature units inside the target to obtain the number of pixels in the printed pattern. The calculation module is used to calculate the occlusion rate based on the total number of pixels and the number of pixels in the printed graphic; The target grayscale image includes a front target grayscale image and a back target grayscale image, and the occlusion rate includes a front occlusion rate and a back occlusion rate. The step of calculating the occlusion rate based on the total number of pixels and the number of pixels in the printed graphic includes: According to formula F shading = (a+b+c) / A, calculate the front shading rate; According to formula R shading = (d+e+f+g) / A, calculate the back shading rate; Among them, F shading Let be the front occlusion rate, 'a' be the number of pixels in the front target border, 'b' be the number of pixels in the front target main grid, 'c' be the number of pixels in the front target fine grid, 'A' be the total number of pixels, and 'R' be the total number of pixels. shading denoted as the back-side occlusion rate, d as the number of pixels in the back-side target border, e as the number of pixels in the back-side target main gate, f as the number of pixels in the back-side target fine gate, and g as the number of pixels in the back-side target electrode.

6. The testing system according to claim 5, characterized in that, The image processing module includes an enhancement and sharpening unit and a filtering and noise reduction unit; The enhancement and sharpening unit is used to enhance and sharpen the target grayscale image; The filtering and noise reduction unit is used to perform filtering and noise reduction processing on the enhanced and sharpened grayscale image.

7. The testing system according to claim 5, characterized in that, The directional gradient histogram conversion module includes a gradient calculation unit, a segmentation unit, and a gradient histogram construction unit; The gradient calculation unit is used to calculate the gradient of the reference grayscale image or the target grayscale histogram; The segmentation unit is used to segment the reference grayscale image or the target grayscale histogram to form several image units; The gradient histogram construction unit is used to calculate the gradient histogram of each image unit based on the magnitude and direction of the gradient to obtain the reference gradient histogram or the target gradient histogram.

8. The testing system according to claim 5, characterized in that, The overlap comparison module includes a comparison unit, a feature calculation unit, and a feature extraction unit; The comparison unit is used to compare the benchmark gradient histogram and the target gradient histogram and remove the overlapping gradient histograms to obtain the target feature gradient histogram. The feature calculation unit is used to calculate the target features based on the target feature gradient histogram; The feature extraction unit is used to obtain all RGB values ​​of the target feature, and to distinguish the target edge and each target internal feature unit in the target grayscale image according to the corresponding RGB values.

9. A computer device, comprising a memory and a processor, characterized in that, The memory stores a computer program, characterized in that the processor executes the computer program to implement the steps of the method according to any one of claims 1 to 4.

Images