A method for testing the silver paste coating specifications of oxide LCD modules

Abstract

This invention relates to the field of LCD module testing technology, and in particular to a method for detecting the silver paste coating specifications of oxide LCD modules. The method includes the following steps: image acquisition: scanning the oxide LCD module line by line on a conveyor belt on an assembly line, capturing each pixel of each line of the moving oxide LCD module and converting them into digital signals; obtaining the Mark points of the oxide LCD module; obtaining the size of the object in the image; establishing a distance reference coordinate system; using this coordinate system to measure and analyze the silver paste specifications; determining the position and size of each point on the target object by measuring the distance and direction from the origin of the coordinate system; distinguishing the silver paste coating areas; and determining the coating specifications. This invention can quickly detect the silver paste specifications of oxide LCD modules after silver paste coating in real time, which helps to improve the production efficiency of module production lines, reduce production costs, and realize a fully automated module industry.

Description

Technical Field

[0001] This invention relates to the field of LCD module testing technology, and in particular to a method for testing the silver paste coating specifications of oxide LCD modules. Background Technology

[0002] Currently, LCD display modules are evolving rapidly, with new technologies and processes constantly being updated and iterated, leading to increasingly widespread applications. Due to the excellent performance and stability of oxide LCDs, they are beginning to be used extensively in fields such as TCP and automotive. Therefore, the requirements for their process inspection are also higher. Silver paste coating is an essential and crucial process for oxide LCD modules, playing a vital role in their display performance. Currently, the main processes of oxide LCD modules are automated; however, existing silver paste inspection generally uses manual microscopic inspection, which suffers from low efficiency, missed or incorrect dimensional and specification checks, and other problems. This hinders the development of intelligent manufacturing module factories and severely limits the application and technological advancement of oxide LCD modules. The silver paste inspection of oxide LCD modules has the following shortcomings:

[0003] 1. Slow testing speed: Traditional testing methods require manual operation, which is time-consuming and labor-intensive, and the testing speed is slow, which cannot meet the needs of the production line.

[0004] 2. Poor reliability: Due to human factors and equipment limitations, existing detection technologies have certain defects in terms of reliability, and are prone to missed detections or false detections.

[0005] 3. High cost: Traditional testing methods require specialized equipment and manpower, and each person needs to be equipped with a dedicated microscopic examination device, which is costly and not conducive to cost control for enterprises. Summary of the Invention

[0006] This invention addresses the problems of existing technologies by providing a method for detecting the specifications of silver paste coating in oxide LCD modules. The method is ingeniously designed and can quickly detect the specifications of silver paste in oxide LCD modules after coating, which helps to improve the production efficiency of module production lines, reduce production costs, and realize a fully automated module industry.

[0007] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0008] This invention provides a method for detecting the specifications of silver paste coating in oxide LCD modules, comprising the following steps:

[0009] Step S10, Image Acquisition: The oxide LCD module is scanned line by line on the conveyor belt of the production line. The pixels of each line of the moving oxide LCD module are captured one by one and converted into digital signals.

[0010] Step S20: Based on SIFT, identify the Mark points of the oxide LCD module;

[0011] Step S30, Pixel size representation: Obtain the size of the object in the image;

[0012] Step S40: Establish a distance reference coordinate system: The center pixel of the Mark point is determined and used as the origin. The X-axis and Y-axis are established, and the direction of the coordinate axes is determined. After establishing a two-dimensional rectangular coordinate system, the silver medal specifications are measured and analyzed using this coordinate system. The specific position and size of these points are determined by measuring the distance and direction of each point on the target object from the origin of the coordinate system.

[0013] Step S50, Image Processing: Differentiate the silver paste coating area;

[0014] Step S60: Determine the specifications of the silver paste coating.

[0015] The image acquisition method in step S10 is as follows:

[0016] Step S11, Equipment Selection and Parameter Adjustment: Before image acquisition, the parameters of the line scan camera need to be set, including exposure time, gain, and synchronization signal; adjust the position, angle, and conveyor belt speed of the camera to make it face the object and scan at a constant speed.

[0017] Step S12, Image Output: After setting the parameters and completing focus and positioning, start the line scan camera and conveyor belt. The line scan camera performs real-time scanning and image acquisition, and transmits the acquired data to the computer.

[0018] The SIFT-based Mark point recognition method in step S20 is as follows:

[0019] Step S21, Feature extraction of Mark points: In Mark point recognition, key points and descriptors are first extracted from the image of the target object and the reference image, and then the position of the target object in the image is determined by matching the descriptors;

[0020] Step S22, Keypoint Matching: The keypoint matching process uses the nearest neighbor matching algorithm. For each keypoint, the closest keypoint in the reference image is found, and the degree of matching is evaluated by comparing the distance between their descriptors. The matching method is based on a ratio test of nearest neighbor matching. For each keypoint, its nearest and second nearest neighbors are found. If the distance ratio between the nearest and second nearest neighbors is less than a threshold, the match is considered successful; otherwise, the match is considered unsuccessful. The distance ratio is calculated as follows:

[0021]

[0022] Where d1 is the distance between the nearest neighbor descriptor and the current keypoint descriptor, and d2 is the distance between the second nearest neighbor descriptor and the current keypoint descriptor. If the distance ratio is less than a set threshold, the match is considered successful.

[0023] The method for representing pixel size in step S30 is as follows: the size of the object in the image can be obtained by calculating the number of pixels in the image. The calculation method is as follows:

[0024] S = L × N;

[0025] Where S is the size of the object, L is the actual size represented by each pixel, and N is the number of pixels occupied by the object.

[0026] The image processing method in step S50 is as follows:

[0027] Step S51: The method for presetting the image ROI region is as follows: Based on the position of the Mark point of the oxide LCD module and the position of the silver paste coating, the retention thresholds y for the top, bottom, left, and right sides of the image are manually preset respectively. 上 y 下 x 左 x 右 , Where x and y are the pixel coordinates of the pixel in the coordinate system established in step S40;

[0028] Based on the threshold values ​​set above, only those conforming to the formula are retained: The pixels;

[0029] Step S52: Binarization of the ROI region: Convert the ROI image into a grayscale image, and then use Otsu's method to segment it into silver paste regions and other regions; by finding a threshold that maximizes the inter-class variance between the foreground and background, the image is binarized; the specific steps are as follows:

[0030] First, calculate the grayscale histogram of the image and normalize it;

[0031] Then, select a threshold T from gray levels 0 to 255 to divide the image into two parts: pixels less than or equal to T are in one class, and pixels greater than T are in another class.

[0032] Calculate the number of pixels, average gray value, and inter-class variance for the two classes;

[0033] According to the formula for calculating inter-class variance σ 2 =ω1(μ1-μ T ) 2 +ω2(μ2-μ T ) 2 Find the threshold T that maximizes the inter-class variance;

[0034] Where ω1 and ω2 represent the pixel proportions of the two classes, and μ1 and μ2 represent the average gray values ​​of the two classes, respectively. T This represents the average grayscale value of the entire image.

[0035] By calculating the inter-class variance at different thresholds, the threshold that maximizes the inter-class variance is selected as the final binarization threshold.

[0036] The method for determining the silver paste coating specifications in step S60 is as follows:

[0037] S61. Calculate the number of silver paste coating areas:

[0038] Fine noise in the image is filtered out using opening and closing operations. Then, the number of silver paste regions, n, is calculated using the contour lookup algorithm in OpenCV. Finally, the number of silver paste regions, n, is compared with the set number of coating regions, n0. set Compare, if n≠n set If the oxide LCD module has a break or missing coating during silver paste application, it is directly determined to be non-compliant with specifications; otherwise, the process proceeds to step S62 after this determination.

[0039] S62: Calculate the width and height of the silver paste coating:

[0040] First, the extreme values ​​x and y of each silver paste coating region are obtained one by one using traversal and sorting algorithms. min x max y min y max The calculation method is as follows:

[0041]

[0042] Where MAX and MIN are functions for finding the maximum and minimum values, respectively;

[0043] Then, the number of pixels in the height and width of the silver paste coating area is obtained by using the extreme values ​​of the two coordinate axes. Finally, the distance dimension S represented by each pixel is used to determine the number of pixels. pixel The height h and width w of the silver paste are calculated as follows:

[0044]

[0045] Finally, the height h and width w of the silver paste are compared with the set thresholds for the height and width of the silver paste coating. If h and w conform to the formula: If the height and width of the silver paste coating are both within the threshold range, then the specifications of the silver paste coating can be determined to meet the requirements; otherwise, they do not meet the requirements.

[0046] The beneficial effects of this invention are:

[0047] This invention is ingeniously designed, using the Mark points of oxide LCDs as a reference benchmark to establish a reference coordinate system. It uses image processing algorithms and mathematical principles to determine the specifications of silver paste coating. This invention achieves rapid and real-time determination of silver paste specifications while ensuring the accuracy of the determination, greatly improving the production efficiency of oxide LCD modules. It can quickly detect the silver paste specifications of oxide LCD modules after silver paste coating in real time, which helps to improve the production efficiency of module production lines, reduce production costs, and realize a fully automated module industry. Attached Figure Description

[0048] Figure 1 This is a hardware layout diagram of a silver paste coating specification detection method for an oxide LCD module according to the present invention.

[0049] Figure 2 This is a flowchart of a method for detecting the specifications of silver paste coating in an oxide LCD module according to the present invention.

[0050] Figure 3 A schematic diagram illustrating the principle of establishing a distance reference coordinate system for determining the specifications of silver paste in this invention.

[0051] Figure 4 This is a schematic diagram illustrating the steps and process for determining the specifications of silver paste according to the present invention. Detailed Implementation

[0052] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to embodiments and accompanying drawings. The content mentioned in the embodiments is not intended to limit the present invention. The present invention will be described in detail below with reference to the accompanying drawings.

[0053] This application provides a method for detecting the specifications of silver paste coating in an oxide LCD module, which includes the following steps:

[0054] Step S10, Image Acquisition: The oxide LCD module is scanned line by line on the conveyor belt of the production line. The pixels of each line of the moving oxide LCD module are captured one by one and converted into digital signals.

[0055] Step S20: Based on SIFT, identify the Mark points of the oxide LCD module;

[0056] Step S30, Pixel size representation: Obtain the size of the object in the image;

[0057] Step S40: Establish a distance reference coordinate system: The center pixel of the Mark point is determined and used as the origin. The X-axis and Y-axis are established, and the direction of the coordinate axes is determined. After establishing a two-dimensional rectangular coordinate system, the silver medal specifications are measured and analyzed using this coordinate system. The specific position and size of these points are determined by measuring the distance and direction of each point on the target object from the origin of the coordinate system.

[0058] Step S50, Image Processing: Differentiate the silver paste coating area;

[0059] Step S60: Coating specification determination.

[0060] Furthermore, the hardware device layout design of this application embodiment is as follows: Figure 1 As shown, it includes the following devices:

[0061] Drive shafts and conveyor belts: As the power source for the flow of oxide LCD modules, their unimpeded operation not only provides continuous and stable power support for the production line, but also effectively promotes smooth flow, thereby ensuring efficient production and high-quality output of LCD modules. Furthermore, this also provides the foundation for imaging by industrial line scan cameras, offering crucial support for production process monitoring and quality control.

[0062] Strip-shaped industrial linear array light sources play a crucial role in the imaging of industrial line scan cameras. By providing a stable and uniform light source, strip-shaped industrial linear array light sources can effectively eliminate shadows and reflections, improving image clarity and accuracy. Furthermore, they help reduce noise and improve image contrast, thus better revealing the details and features of the inspected object. Through proper light source design and arrangement, the imaging quality of industrial line scan cameras can be effectively improved, providing reliable support for production process monitoring and quality control.

[0063] Industrial line scan cameras: Industrial line scan cameras can quickly and accurately acquire image information of the object to be inspected by receiving light reflected from different materials on the surface of an oxide LCD, thus realizing inspection tasks on automated production lines.

[0064] Specifically, under the above settings, the embodiments of this application can quickly detect the silver paste specifications of oxide LCD modules after silver paste coating in real time, which helps to improve the production efficiency of module production lines, reduce production costs, and provide a high-quality solution for the fully automated module industry. In this embodiment, the Mark points of oxide LCD are used as reference benchmarks to establish a reference coordinate system, and image processing algorithms and mathematical principles are used to determine the silver paste coating specifications. The embodiments of this application can flexibly adjust hardware equipment such as camera light sources according to different precision requirements in actual production, realizing rapid and real-time silver paste specification determination while ensuring the accuracy of determination, which greatly improves the production efficiency of oxide LCD modules.

[0065] In this embodiment, image acquisition is performed using a line scan camera, in conjunction with a conveyor belt on an assembly line to scan the oxide LCD module line by line. The moving oxide LCD module captures each pixel of its row individually and converts them into digital signals. The image acquisition method in step S10 is as follows:

[0066] Step S11, Equipment Selection and Parameter Adjustment: Before image acquisition, the parameters of the line scan camera need to be set, including exposure time, gain, and synchronization signal; adjust the position, angle, and conveyor belt speed of the camera to make it face the object and scan at a constant speed.

[0067] Step S12, Image Output: After setting the parameters and completing focus and positioning, start the line scan camera and conveyor belt. The line scan camera performs real-time scanning and image acquisition, and transmits the acquired data to the computer.

[0068] Among them, the Mark point is a special mark in the LCD module; it is usually in the shape of a "+", located at the lower left and lower right corners of the module terminal side. It serves as the basis for the automated equipment of the entire module, used for positioning, calibration, and testing of the module during production and assembly. Therefore, the silver paste specification testing also requires the use of Mark points as the basis for the entire method. The principle of the SIFT-based Mark point recognition method is to use the SIFT algorithm to extract key points and descriptors in the image, and then identify the Mark points of the target object by matching these descriptors. The SIFT-based Mark point recognition method in step S20 is as follows:

[0069] Step S21, Feature extraction of Mark points: In Mark point recognition, key points and descriptors are first extracted from the image of the target object and the reference image, and then the position of the target object in the image is determined by matching the descriptors;

[0070] Step S22, Keypoint Matching: The keypoint matching process uses the nearest neighbor matching algorithm. For each keypoint, the closest keypoint in the reference image is found, and the degree of matching is evaluated by comparing the distance between their descriptors. The matching method is based on a ratio test of nearest neighbor matching. For each keypoint, its nearest and second nearest neighbors are found. If the distance ratio between the nearest and second nearest neighbors is less than a threshold, the match is considered successful; otherwise, the match is considered unsuccessful. The distance ratio is calculated as follows:

[0071]

[0072] Where d1 is the distance between the nearest neighbor descriptor and the current keypoint descriptor, and d2 is the distance between the second nearest neighbor descriptor and the current keypoint descriptor. If the distance ratio is less than a set threshold, the match is considered successful.

[0073] The method for representing pixel size in step S30 is as follows: the size of the object in the image can be obtained by calculating the number of pixels in the image. The calculation method is as follows:

[0074] S = L × N;

[0075] Where S is the size of the object, L is the actual size represented by each pixel, and N is the number of pixels occupied by the object.

[0076] In this embodiment, the surface of the glass in the oxide LCD module undergoes specular reflection under the illumination of a linear array light source, while the surface of the silver paste undergoes diffuse reflection. Therefore, there is a significant difference between the surfaces of the silver paste and other materials. The brightness of the glass surface in the oxide LCD module is significantly higher than that of the silver paste surface. Therefore, image processing algorithms can be used to distinguish the silver paste coating area. The image processing method in step S50 is as follows:

[0077] Step S51: The image acquired by the line scan camera has high resolution and contains a large number of redundant pixels. To improve the processing speed of the algorithm, redundant pixels in the image can be significantly reduced by presetting the Region of Interest (ROI). By setting the ROI, the number of unnecessary pixels in the image can be effectively reduced, thereby reducing the amount of data and improving the processing efficiency of the algorithm. The method for presetting the image ROI is as follows: based on the position of the Mark point of the oxide LCD module and the position of the silver paste coating, the retention thresholds y for the top, bottom, left, and right sides of the image are manually preset. 上 y 下 x 左 x 右 , Where x and y are the pixel coordinates of the pixel in the coordinate system established in step S40;

[0078] Based on the threshold values ​​set above, only those conforming to the formula are retained: The pixels;

[0079] Step S52: Binarization of the ROI region: Convert the ROI image into a grayscale image, and then use Otsu's method to segment it into silver paste regions and other regions; by finding a threshold that maximizes the inter-class variance between the foreground and background, the image is binarized; the specific steps are as follows:

[0080] First, calculate the grayscale histogram of the image and normalize it;

[0081] Then, select a threshold T from gray levels 0 to 255 to divide the image into two parts: pixels less than or equal to T are in one class, and pixels greater than T are in another class.

[0082] Calculate the number of pixels, average gray value, and inter-class variance for the two classes;

[0083] According to the formula for calculating inter-class variance σ 2 =ω1(μ1-μ T ) 2 +ω2(μ2-μ T ) 2 Find the threshold T that maximizes the inter-class variance;

[0084] Where ω1 and ω2 represent the pixel proportions of the two classes, and μ1 and μ2 represent the average gray values ​​of the two classes, respectively. T This represents the average grayscale value of the entire image.

[0085] By calculating the inter-class variance under different thresholds, the threshold that maximizes the inter-class variance is selected as the final binarization threshold, thus achieving Otsu's binarization method. This method can effectively and adaptively select the optimal threshold based on the characteristics of the image, and is suitable for image segmentation of silver paste and glass in oxide LCDs.

[0086] In this embodiment, according to the above process, the area to be coated with silver paste has been successfully determined, and the conditions for determining the coating specifications have been met. Now, it is only necessary to calculate the relevant parameters using the existing conditions and compare them with the set silver paste specification parameters to determine the specifications of the silver paste coating in real time. The method for determining the coating specifications in step S60 is as follows:

[0087] S61. Calculate the number of silver paste coating areas:

[0088] First, opening and closing operations are used to filter out minor noise in the image. Then, the number of silver paste regions, n, is calculated using the contour lookup algorithm in OpenCV. Finally, the number of silver paste regions, n, is compared with the set number of coating regions, n'. set Compare, if n≠n setIf the oxide LCD module has a break or missing coating during silver paste application, it is directly determined to be non-compliant with specifications; otherwise, the process proceeds to step S62 after this determination.

[0089] S62: Calculate the width and height of the silver paste coating:

[0090] First, the extreme values ​​x and y of each silver paste coating region are obtained one by one using traversal and sorting algorithms. min x max y min y max The calculation method is as follows:

[0091]

[0092] Where MAX and MIN are functions for finding the maximum and minimum values, respectively;

[0093] Then, the number of pixels in the height and width of the silver paste coating area is obtained by using the extreme values ​​of the two coordinate axes. Finally, the distance dimension S represented by each pixel is used to determine the number of pixels. pixel The height h and width w of the silver paste are calculated as follows:

[0094]

[0095] Finally, the height h and width w of the silver paste are compared with the set thresholds for the height and width of the silver paste coating. If h and w conform to the formula: If the height and width of the silver paste coating are both within the threshold range, then the specifications of the silver paste coating can be determined to meet the requirements; otherwise, they do not meet the requirements.

[0096] Key point 1 of this application embodiment: A production line acquisition system for oxide LCD modules using a line scan camera was built; Key point 2: The SIFT algorithm was used to identify the Mark points of the oxide LCD, and a two-axis coordinate system was established based on the Mark points as a specification judgment benchmark; Key point 3: Pixels are used to represent the actual size, which allows the present invention to flexibly adjust hardware devices such as cameras, lenses and light sources according to actual accuracy requirements, which is beneficial to the production cost of oxide LCD modules.

[0097] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes, and modifications made to the above embodiments based on the present invention without departing from the scope of the present invention are within the scope of the present invention.

Claims

1. A method for detecting the specifications of silver paste coating in an oxide LCD module, characterized in that, Includes the following steps: Step S10, Image Acquisition: The oxide LCD module is scanned line by line on the conveyor belt of the production line. The pixels of each line of the moving oxide LCD module are captured one by one and converted into digital signals. Step S20: Based on SIFT, identify the Mark points of the oxide LCD module; Step S30, Pixel size representation: Obtain the size of the object in the image; Step S40: Establish a distance reference coordinate system: The center pixel of the Mark point is determined and used as the origin. The X-axis and Y-axis are established, and the direction of the coordinate axes is determined. After establishing a two-dimensional rectangular coordinate system, the silver paste specifications are measured and analyzed using this coordinate system. The specific position and size of these points are determined by measuring the distance and direction of each point on the target object from the origin of the coordinate system. Step S50, Image Processing: Differentiate the silver paste coating area; Step S60: Determining the specifications of the silver paste coating; The SIFT-based Mark point recognition method in step S20 is as follows: Step S21, Feature extraction of Mark points: In Mark point recognition, key points and descriptors are first extracted from the image of the target object and the reference image, and then the position of the target object in the image is determined by matching the descriptors; Step S22, Keypoint Matching: The keypoint matching process uses the nearest neighbor matching algorithm. For each keypoint, the closest keypoint in the reference image is found, and the degree of matching is evaluated by comparing the distance between their descriptors. The matching method is based on a ratio test of nearest neighbor matching. For each keypoint, its nearest and second nearest neighbors are found. If the distance ratio between the nearest and second nearest neighbors is less than a threshold, the match is considered successful; otherwise, the match is considered unsuccessful. The distance ratio is calculated as follows: ； in, d 1 represents the distance between the nearest neighbor descriptor and the current keypoint descriptor. d 2 is the distance between the next nearest neighbor descriptor and the current keypoint descriptor. If the distance ratio is less than a set threshold, the match is considered successful. The method for determining the silver paste coating specifications in step S60 is as follows: S61. Calculate the number of silver paste coating areas: Fine noise in the image is filtered out using opening and closing operations. Then, the number of silver paste regions is calculated using the contour lookup algorithm in OpenCV. n Finally, compare the number of silver paste regions. n and the set number of coating areas n set If a comparison is made, n ≠ n set If the oxide LCD module has a break or missing coating during silver paste application, it is directly determined to be non-compliant with specifications; otherwise, the process proceeds to step S62 after this determination. S62: Calculate the width and height of the silver paste coating: First, each silver paste coating area is calculated one by one using traversal and sorting algorithms. x shaft and y Extreme values ​​of the axis x min , x max , y min , y max The calculation method is as follows: ； in, MAX and MIN These are functions for finding the maximum and minimum values, respectively; Then, the number of pixels in the height and width of the silver paste coating area is obtained by using the extreme values ​​of the two coordinate axes. Finally, the distance dimension represented by each pixel is used. S pixel Calculate the height of the silver paste h Hekuan w The calculation method is as follows: ； Finally, the high-quality silver paste h Hekuan w Compared with the set thresholds for the height and width of the silver paste coating, if h and w Conforms to the formula: If the height and width of the silver paste coating are both within the threshold range, then the specifications of the silver paste coating can be determined to meet the requirements; otherwise, they do not meet the requirements.

2. The method for detecting the silver paste coating specifications of an oxide LCD module according to claim 1, characterized in that, The image acquisition method in step S10 is as follows: Step S11, Equipment Selection and Parameter Adjustment: Before image acquisition, the parameters of the line scan camera need to be set, including exposure time, gain, and synchronization signal; adjust the position, angle, and conveyor belt speed of the camera to make it face the object and scan at a constant speed. Step S12, Image Output: After setting the parameters and completing focus and positioning, start the line scan camera and conveyor belt. The line scan camera performs real-time scanning and image acquisition, and transmits the acquired data to the computer.

3. The method for detecting the silver paste coating specifications of an oxide LCD module according to claim 1, characterized in that, The method for representing pixel size in step S30 is as follows: the size of the object in the image can be obtained by calculating the number of pixels in the image. The calculation method is as follows: ； in, S The size of the object, L The actual size represented by each pixel. N This represents the number of pixels occupied by the object.

4. The method for detecting the silver paste coating specifications of an oxide LCD module according to claim 1, characterized in that, The image processing method in step S50 is as follows: Step S51: The method for presetting the image ROI region is as follows: Based on the position of the Mark point of the oxide LCD module and the position of the silver paste coating, the retention thresholds for the top, bottom, left, and right sides of the image are manually preset respectively. y 上 , y 下 , x 左 , x 右 , ; in, x , y The pixel coordinates are the coordinates of the pixel in the coordinate system established in step S40; Based on the threshold values ​​set above, only those conforming to the formula are retained: The pixels; Step S52: Binarization of the ROI region: Convert the ROI image into a grayscale image, and then use Otsu's method to segment it into silver paste regions and other regions; by finding a threshold that maximizes the inter-class variance between the foreground and background, the image is binarized; the specific steps are as follows: First, calculate the grayscale histogram of the image and normalize it; Then, select a threshold T from gray levels 0 to 255 to divide the image into two parts: pixels less than or equal to T are in one class, and pixels greater than T are in another class. Calculate the number of pixels, average gray value, and inter-class variance for the two classes; According to the formula for calculating inter-class variance Find the threshold T that maximizes the inter-class variance; Among them, respectively and The pixel percentage of the two classes. and These are the average gray values ​​of the two classes, This represents the average grayscale value of the entire image. By calculating the inter-class variance at different thresholds, the threshold that maximizes the inter-class variance is selected as the final binarization threshold.

Images