LED circuit board defect identification method based on optical imaging

Abstract

The invention discloses an LED circuit board defect identification method based on optical imaging, and relates to the technical field of defect identification, and the method comprises the steps: carrying out the imaging capture of an LED circuit board through an optical imaging device, obtaining an optical image of the LED circuit board, carrying out the image processing of the optical image, generating a corresponding standard pattern optical image, and carrying out the recognition of the defect of the LED circuit board. And performing feature extraction on the standard pattern optical graph to obtain corresponding image optical feature information, constructing a defect identification model, inputting the image optical feature information into the defect identification model, and further classifying and identifying the defect type corresponding to the LED circuit board. The classification and recognition result of the defect recognition model on the LED circuit board is fed back to an operation terminal corresponding to the production line, the operation terminal controls a mechanical arm arranged on the production line, and then the abnormal LED circuit board is sorted out.

Description

Technical Field

[0001] The present invention relates to the technical field of defect recognition, and in particular to a method for identifying defects in LED circuit boards based on optical imaging. Background Art

[0002] With the rapid development of LED technology, LED circuit boards are increasingly used in lighting, display screens and other fields. However, during the production process, LED circuit boards may have various defects, such as solder joint defects, circuit breakage, electrical connection defects, etc. These defects seriously affect the quality and service life of the product.

[0003] In addition, the traditional defect detection method mainly relies on manual visual inspection after the production of LED circuit boards is completed, which is inefficient and has low accuracy. It cannot promptly detect abnormal LED circuit boards on the production line and screen them out in time. Therefore, it is of great significance to study an efficient and accurate LED circuit board defect recognition method to promptly detect defective LED circuit boards on the production line and ensure the production quality of LED circuit boards. Summary of the Invention

[0004] In order to solve the above problems, the purpose of the present invention is to provide a method for identifying defects of LED circuit boards based on optical imaging.

[0005] The purpose of the present invention can be achieved by the following technical solution: A method for identifying defects in LED circuit boards based on optical imaging, comprising the following steps: Step S1: capturing an LED circuit board through an optical imaging device to obtain an optical image of the LED circuit board; Step S2: performing image processing on the optical image to generate a corresponding standard optical image, performing feature extraction on the standard optical image to obtain corresponding image optical feature information; Step S3: Construct a defect recognition model and input the image optical feature information into the defect recognition model to classify and identify the defect type corresponding to the LED circuit board; Step S4: Feedback the classification and identification results of the defect recognition model on the LED circuit boards to the operation terminal corresponding to the production line, and the operation terminal controls the robotic arm arranged on the production line to sort out abnormal LED circuit boards.

[0006] Furthermore, the process of capturing the LED circuit board by an optical imaging device and obtaining an optical image of the LED circuit board includes: Set device parameters of optical imaging equipment; For each processing stage of the LED circuit board, a corresponding circuit board storage area is set up on the production line. The optical imaging equipment is arranged in the circuit board storage area on the production line. The circuit board storage area is used to store LED circuit boards. Start the optical imaging device to capture the LED circuit board and obtain optical images of the LED circuit board at several angles. When the imaging area captured by the imaging cannot cover the board area of ​​the LED circuit board, adjust the layout coordinates of the optical imaging device on the production line to perform regional shooting, and splice the regional optical images obtained by the regional shooting to obtain a complete optical image of the LED circuit board.

[0007] Furthermore, the process of performing image processing on the optical image to generate a corresponding standard style optical image includes: Determine whether the optical image is distorted and the type of distortion corresponding to the distortion; Perform corresponding distortion correction processing on the optical image according to the distortion type; Setting a pixel cell window, dividing the optical image into a plurality of image sub-regions through the pixel cell window, and obtaining the regional image brightness and regional image contrast of each image sub-region; Determine whether the regional image brightness and regional image contrast of the image sub-region meet the requirements, and process the image sub-region that does not meet the requirements; Each optical image is cropped to a preset standard size to generate a standard style optical map.

[0008] Furthermore, the process of determining whether the regional image brightness and the regional image contrast of the image sub-region meet the requirements and processing the image sub-region that does not meet the requirements includes: The image subregions are labeled 1, 2, 3, ..., k, where k is a natural number greater than 0, tS[k] is the regional image brightness of the image subregion numbered k, Db[k] is the regional image contrast of the image subregion numbered k, and the brightness threshold and contrast threshold are set and recorded as D 亮度 and D 对比度 ; When tS[k]≥D 亮度 When tS[k]<D 亮度 When , no operation is performed; When Db[k]≥D 对比度 When Db[k]<D 对比度 , no action is performed.

[0009] Furthermore, the process of extracting features from the standard optical image to obtain corresponding image optical feature information includes: Detect image key points in standard style optical images using SIFT algorithm; Using the Canny edge detection algorithm to identify the edge contours of the LED circuit board and the edge contours of the distributed circuits on the LED circuit board in the standard optical pattern, and combining them as the edge contour information of the standard optical pattern; Feature extraction of the standard optical pattern includes shape feature extraction, texture feature extraction, color feature extraction, and frequency domain feature extraction, thereby obtaining the geometric properties, texture information, color distribution information, and frequency domain characteristics of the LED circuit board and the circuits distributed on the LED circuit board corresponding to the current standard optical pattern; The image key points, edge contour information, geometric properties, texture information, color distribution information and frequency domain features corresponding to the standard style optical image are integrated and then used as the image optical feature information of the corresponding standard style optical image.

[0010] Furthermore, the process of building a defect recognition model includes: Construct the initial CNN model through convolutional neural network technology; Select LED circuit boards that have been classified into defect types and LED circuit boards without defects, obtain the corresponding optical feature information of their respective standard style optical images after processing, and build the information pooling layer and data output layer of the CNN model; The optical feature information is input into the information pooling layer to obtain the corresponding main optical features and the pooling coefficients corresponding to the main optical features. The part with the highest data proportion in the optical feature information is used as the main optical feature, and the defect judgment interval is set; Based on the subordinate relationship between the pooling coefficient of the main optical features and the defect judgment interval, it is determined whether the currently identified LED circuit board has defects and the type of defects identified. The recognition result is output by the data output layer, and the recognition result is compared with the actual result. Based on the comparison result, it is decided whether to train the CNN model, and then the final defect recognition model is constructed.

[0011] Furthermore, the process of inputting the image optical feature information into the defect recognition model and then classifying and identifying the defect type corresponding to the LED circuit board includes: When the defect recognition model is successfully constructed, the image optical feature information corresponding to the currently unclassified LED circuit board is input into the defect recognition model to classify and identify the defect types of the LED circuit board, including solder joint defects, electrical connection defects, circuit defects, and interlayer defects.

[0012] Furthermore, the classification and identification results of the LED circuit boards by the defect recognition model are fed back to the operation terminal corresponding to the production line. The operation terminal controls the robotic arm arranged on the production line to sort out abnormal LED circuit boards. The process includes: The defect recognition model feeds back the real-time recognition results to the corresponding operation terminal of the production line, which then locates the defective LED circuit board and the corresponding circuit board storage area on the production line. Each circuit board storage area is equipped with a corresponding robotic arm. When the identification result is received that the LED circuit board has a defect, the corresponding LED circuit board is marked as an abnormal LED circuit board. The robotic arm is controlled to pick out the abnormal LED circuit boards in each circuit board storage area and classify and store them according to the defect type.

[0013] Compared with the prior art, the beneficial effects of the present invention are: obtaining an optical image of an LED circuit board through an optical imaging device, generating a corresponding standard style optical image after image processing, and obtaining corresponding image optical feature information after feature extraction, constructing a defect recognition model to input the image optical feature information to classify and identify the defect type corresponding to the LED circuit board, and feeding back the classification and recognition results to the operation terminal corresponding to the production line in real time. The operation terminal controls the robotic arm arranged on the production line, and then picks out abnormal LED circuit boards, effectively improving the efficiency and accuracy of defect identification of LED circuit boards, so as to timely screen out defective LED circuit boards, ensure the processing quality of LED circuit boards, and maintain the efficient operation of the LED circuit board production line. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 Flowchart of the present invention. DETAILED DESCRIPTION

[0015] like Figure 1 As shown, a method for identifying defects in LED circuit boards based on optical imaging includes the following steps: Step S1: capturing an LED circuit board through an optical imaging device to obtain an optical image of the LED circuit board; Step S2: performing image processing on the optical image to generate a corresponding standard optical image, performing feature extraction on the standard optical image to obtain corresponding image optical feature information; Step S3: Construct a defect recognition model and input the image optical feature information into the defect recognition model to classify and identify the defect type corresponding to the LED circuit board; Step S4: Feedback the classification and identification results of the defect recognition model on the LED circuit boards to the operation terminal corresponding to the production line, and the operation terminal controls the robotic arm arranged on the production line to sort out abnormal LED circuit boards.

[0016] It should be further explained that, in a specific implementation process, the process of capturing the LED circuit board by an optical imaging device and then obtaining an optical image of the LED circuit board includes: Set the device parameters of the optical imaging device, including lens focal length, exposure time, and light source illumination; The optical imaging device is arranged in a circuit board storage area on a production line, and the production line is used to process LED circuit boards. Each processing stage of the LED circuit boards on the production line is provided with a corresponding circuit board storage area on the production line; The circuit board storage area is used to store LED circuit boards; The LED circuit boards stored in each circuit board storage area are marked with the corresponding circuit board code and storage area code, which are used as the unique authentication identifier of the LED circuit board at the current location; When the LED circuit boards on the production line enter the circuit board storage area, the optical imaging device is started and the LED circuit boards are imaged and captured according to the set equipment parameters, thereby obtaining optical images of the LED circuit boards at several angles to cover the entire surface area of ​​the LED circuit boards; When the imaging area captured by the optical imaging device cannot cover the board area of ​​the LED circuit board, the layout coordinates of the optical imaging device on the production line are adjusted, the LED circuit board is photographed in different areas, and the regional optical images obtained by the regional photography are spliced ​​to obtain a complete optical image of the LED circuit board.

[0017] It should be further explained that, in a specific implementation process, the process of performing image processing on an optical image and generating a corresponding standard optical image includes: Determine whether the optical image is distorted and the type of distortion corresponding to the distortion; And perform corresponding distortion correction processing on the optical image according to the distortion type; If the optical image is distorted and the distortion type is radial distortion, a distortion model is constructed; Obtain distortion parameters, including barrel distortion parameters and pincushion distortion parameters; The barrel distortion parameter corresponds to the barrel distortion in radial distortion, which is manifested as the straight line in the optical image bulging outward. The barrel distortion parameter is a positive value and is denoted as p1, p2, p3, ..., pn, where n is an integer greater than 1. Among them, the order of p1-pn gradually increases, p1 is the main control parameter, and p2-pn is used to describe the distortion details in barrel distortion; The pincushion distortion parameter corresponds to the pincushion distortion in the radial distortion, which is manifested as the straight line in the optical image being concave inward. The pincushion distortion parameter is a negative value and is denoted as q1, q2, q3, ..., qm, where m is an integer greater than 1. Among them, the order of q1-qm gradually increases, q1 is the main control parameter, and q2-qm is used to describe the distortion details in the pincushion distortion; Taking the barrel distortion in radial distortion as an example, it can be expressed as follows: X-distorted=X(1+p1×r 2 +p2×r 2 +p3×r 2 +……pn×r 2 ); Y-distorted=Y(1+p1×r 2 +p2×r 2 +p3×r 2 +……pn×r 2 ); Where (X, Y) is the coordinate of the point in the ideal undistorted optical image, (X-distorted, Y-distorted) is the coordinate of the point in the distorted optical image, and r 2 =X 2 +Y 2 is the square of the distance from the point to the center of the optical image; The distortion parameters of the optical image are input into the distortion model, and the ideal position of each pixel in the optical image is calculated by the distortion model. Then, the distortion of the optical image is corrected, and the pixel value of each pixel after correction is calculated using bilinear interpolation. When the distortion type is perspective distortion, calibrate the four corner points in the optical image, define their ideal positions in the optical image, and correct the four corner points using the perspective transformation matrix to eliminate the perspective distortion. When the distortion type is chromatic aberration, the color balance of the optical image is adjusted using image processing software, and color deviations are automatically corrected using an algorithm.

[0018] Set the pixel cell window, which consists of 6×6 pixel cells; Dividing the optical image into a number of image sub-regions through a pixel cell window; Obtain the unit brightness and unit contrast of each pixel cell in each image sub-region, and accumulate the unit brightness and unit contrast of all pixel cells in each image sub-region to generate the regional image brightness and regional image contrast of each image sub-region; The image sub-regions are labeled 1, 2, 3, ..., k, where k is a natural number greater than 0; tS[k] is the regional image brightness of the image subregion numbered k; Db[k] is the regional image contrast of the image subregion numbered k; Set the brightness threshold and contrast threshold, and record them as D 亮度 and D 对比度 ; When tS[k]≥D 亮度 When tS[k]<D 亮度 When , no operation is performed; When Db[k]≥D 对比度 When Db[k]<D 对比度 When , no operation is performed; Each optical image is cropped to a preset standard size to generate a standard style optical map.

[0019] It should be noted that in practical applications, only the first few distortion parameters need to be considered, because the influence of higher-order distortion parameters is smaller and it is more difficult to accurately estimate them through the calibration process. Usually, p1, p2, q1, and q2 are sufficient to describe the radial distortion of most lenses. If higher correction accuracy is required, p3, p4 or higher-order distortion parameters may be considered.

[0020] It should be further explained that, in a specific implementation process, the process of extracting features from a standard optical image and obtaining corresponding image optical feature information includes: The SIFT algorithm is used to detect image key points in standard optical images. Image key points include corner points, edge points, spots, endpoints, T-junctions, and high curvature points. Corner point: The point formed by the intersection of two edges in a standard style optical diagram; Edge point: A point on the edge of a standard optical diagram, used to represent the outline of an LED circuit board; Spot: An area of ​​uniform texture or color in a standard optical pattern, used to represent the location of circuits on an LED circuit board. Endpoint: The starting or ending point of an edge in a standard-style optical diagram; T-junction: The point where one edge segment terminates at another edge segment in a standard optical diagram. Using the Canny edge detection algorithm to identify the edge contours of the LED circuit board and the edge contours of the distributed circuits on the LED circuit board in the standard optical pattern, and combining them as the edge contour information of the standard optical pattern; Feature extraction of the standard optical pattern includes shape feature extraction, texture feature extraction, color feature extraction, and frequency domain feature extraction, thereby obtaining the geometric properties, texture information, color distribution information, and frequency domain characteristics of the LED circuit board and the circuits distributed on the LED circuit board corresponding to the current standard optical pattern; Geometric properties include area, perimeter, rectangularity, and circularity; Texture information is extracted through local binary patterns and gray-level co-occurrence matrix; Color distribution information is obtained by calculating the color histogram and constructing the color moment; The changes of the standard style image in the frequency domain are extracted as frequency domain features through Fourier transform; The image key points, edge contour information, geometric properties, texture information, color distribution information and frequency domain features corresponding to the standard style optical image are integrated and then used as the image optical feature information of the corresponding standard style optical image.

[0021] It should be further explained that, in the specific implementation process, the process of building a defect identification model includes: Construct the initial CNN model through convolutional neural network technology; Select LED circuit boards that have been classified into defect types and LED circuit boards without defects, obtain the corresponding optical feature information of their respective standard style optical images after processing, and build the information pooling layer and data output layer of the CNN model; The optical feature information is input into the information pooling layer to obtain the corresponding main optical features and the corresponding pooling coefficients. The pooling coefficients are used to characterize the overall characteristics of the LED circuit board corresponding to the current standard style optical image, which is used for subsequent defect type identification. The part of the optical feature information with the highest data percentage is taken as the main optical feature and is expressed as follows:

[0022] Among them, D1 is the data volume corresponding to the key points of the image in the optical feature information, D2 is the data volume corresponding to the edge contour information, D3 is the data volume after the accumulation of geometric attributes, texture information, color distribution information and frequency domain features, and D is the data volume corresponding to the entire optical feature information; Key[Switch] is used to select the optical feature information corresponding to the item with the highest value separated by commas in {} as the main optical feature; Set weight coefficients a, b and c; Set the correction coefficients α and β;

[0023] Wherein, a, b and c are all natural numbers greater than 0, 0<α<1, 0<β<1; Set the defect judgment interval and record it as Ω 缺陷 ,Ω 缺陷 =[0.6, 1); When Ch∈Ω 缺陷 When the LED circuit board is identified, it is determined that there is a defect, and the data output layer outputs the identified defect type; It should be noted that when determining that an LED circuit board has defects, further detailed defect types are carried out through automatic optical inspection, X-ray inspection, flying probe testing and functional testing, and the identified defect types are output through the defect recognition model. The defect recognition model uses the defect type of the defective LED circuit board and the relationship between the corresponding optical feature information as the identification reference data for the subsequent defect recognition model, so that the defect recognition model can automatically identify the defect type of the LED circuit board in the future.

[0024] When Ch∉Ω 缺陷 When the LED circuit board is identified, it is determined that there is no defect, and the data output layer outputs the identification result; The recognition results are compared with the actual results of LED circuit boards that have been classified into defect types and LED circuit boards without defects to obtain the model recognition accuracy. When the model recognition accuracy is lower than the preset accuracy threshold, the optical feature information of LED circuit boards that have been classified into defect types and LED circuit boards without defects are added as a training set. The CNN model is trained using the training set, and the weight coefficient and correction coefficient are changed until the model recognition accuracy is not lower than the accuracy threshold. The model training is stopped to construct the final defect recognition model.

[0025] It should be noted that the defect judgment interval can be changed according to the actual model verification and identification situation, and the weight coefficient, correction coefficient and accuracy threshold can all be changed and adjusted according to the actual model verification situation and classification and identification requirements.

[0026] It should be further explained that, in a specific implementation process, the process of inputting the image optical feature information into the defect recognition model and then classifying and identifying the corresponding defect type of the LED circuit board includes: Once the defect recognition model is successfully constructed, the optical feature information of the image corresponding to the currently unclassified LED circuit board is input into the defect recognition model to classify and identify the defect type of the current LED circuit board. The defect types include solder joint defects, electrical connection defects, circuit defects, and interlayer defects. Solder joint defects include cold solder, excessive solder, insufficient solder, solder balls, and open circuits; The details are as follows: Cold welding: The solder joint is not completely melted, resulting in a weak weld; Too much or too little solder: affects electrical connection and mechanical strength; Solder balls: Excess solder balls formed during the soldering process may cause short circuits; Open circuit: The solder joint is not formed, resulting in a disconnected circuit; Electrical connection defects include short circuits and open circuits; The details are as follows: Short circuit: An electrical connection occurs between conductive paths that should not be connected. Open circuit: The conductive path is interrupted, resulting in the inability of current to flow; Circuit defects include board cracks, board holes, and uneven conductor width; The details are as follows: Board cracks: Cracks appear on the copper foil or circuit on the LED circuit board; Holes on the board: The holes on the LED circuit board are not properly punched or the hole walls are defective; Uneven conductor width: Inconsistent conductor width may affect the uniform distribution of current; Interlayer defects include interlayer short circuits and poor interlayer insulation; The details are as follows: Interlayer short circuit: The conductive paths between different layers in a multilayer board are short-circuited; Poor interlayer insulation: The interlayer insulation layer is damaged, resulting in a decrease in electrical performance.

[0027] It should be further explained that, in the specific implementation process, the classification and identification results of the LED circuit boards by the defect recognition model are fed back to the operation terminal corresponding to the production line. The operation terminal controls the robotic arm arranged on the production line to sort out abnormal LED circuit boards. The process includes: The final defect recognition model will feed back the real-time recognition results to the operation terminal corresponding to the production line. The operation terminal will locate the defective LED circuit board and the corresponding circuit board storage area on the production line based on the recognition results, circuit board code and storage area code. Each circuit board storage area is equipped with a corresponding robotic arm. When the identification result is received that the LED circuit board has a defect, the corresponding LED circuit board is marked as an abnormal LED circuit board, and the robotic arm is controlled to pick out the abnormal LED circuit boards in each circuit board storage area and classify and store them according to the defect type of the LED circuit board.

[0028] It should be noted that each circuit board storage area on the production line corresponds to a processing stage of the LED circuit board. Further judgment of whether the LED circuit board in the corresponding circuit storage area is abnormal needs to be made based on the current processing stage and all previous processing operations. For example: there is no soldering operation on the LED circuit board in the current processing stage and all previous processing operations, and the defect type of the LED circuit board in the corresponding circuit board storage area is identified by the defect recognition model as a defect related to the soldering operation, then it cannot be determined that the LED circuit board in the current circuit board storage area is abnormal. However, if there is a defect type corresponding to the current processing stage and all previous processing operations, then it can be determined that the LED circuit board is abnormal. For example: the current processing stage is the lamination stage, and the defect recognition model identifies the defect type related to the lamination stage, then it can be determined that the current LED circuit board is abnormal.

[0029] The above formulas are all calculated by removing dimensions and taking their numerical values. The formula is a formula that is closest to the actual situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and preset thresholds in the formula are set by technicians in this field according to actual conditions or obtained by simulating a large amount of data.

[0030] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A method for identifying defects in LED circuit boards based on optical imaging, characterized in that: The following steps are involved: Step S1: capturing an LED circuit board through an optical imaging device to obtain an optical image of the LED circuit board; Step S2: performing image processing on the optical image to generate a corresponding standard optical image, performing feature extraction on the standard optical image to obtain corresponding image optical feature information; Step S3: Construct a defect recognition model and input the image optical feature information into the defect recognition model to classify and identify the defect type corresponding to the LED circuit board; Step S4: Feedback the classification and identification results of the defect recognition model on the LED circuit boards to the operation terminal corresponding to the production line, and the operation terminal controls the robotic arm arranged on the production line to sort out abnormal LED circuit boards.

2. The method for identifying defects in LED circuit boards based on optical imaging according to claim 1, characterized in that: The process of capturing an LED circuit board through an optical imaging device and obtaining an optical image of the LED circuit board includes: Set device parameters of optical imaging equipment; For each processing stage of the LED circuit board, a corresponding circuit board storage area is set up on the production line. The optical imaging equipment is arranged in the circuit board storage area on the production line. The circuit board storage area is used to store LED circuit boards. Start the optical imaging device to capture the LED circuit board and obtain optical images of the LED circuit board at several angles. When the imaging area captured by the imaging cannot cover the board area of ​​the LED circuit board, adjust the layout coordinates of the optical imaging device on the production line to perform regional shooting, and splice the regional optical images obtained by the regional shooting to obtain a complete optical image of the LED circuit board.

3. The method for identifying defects in LED circuit boards based on optical imaging according to claim 2, characterized in that: The process of processing an optical image to generate a corresponding standard optical image includes: Determine whether the optical image is distorted and the type of distortion corresponding to the distortion; Perform corresponding distortion correction processing on the optical image according to the distortion type; Setting a pixel cell window, dividing the optical image into a plurality of image sub-regions through the pixel cell window, and obtaining the regional image brightness and regional image contrast of each image sub-region; Determine whether the regional image brightness and regional image contrast of the image sub-region meet the requirements, and process the image sub-region that does not meet the requirements; Each optical image is cropped to a preset standard size to generate a standard style optical map.

4. The method for identifying defects in LED circuit boards based on optical imaging according to claim 3, characterized in that: The process of determining whether the regional image brightness and regional image contrast of the image sub-region meet the requirements and processing the image sub-region that does not meet the requirements includes: The image subregions are labeled 1, 2, 3, ..., k, where k is a natural number greater than 0, tS[k] is the regional image brightness of the image subregion numbered k, Db[k] is the regional image contrast of the image subregion numbered k, and the brightness threshold and contrast threshold are set and recorded as D 亮度 and D 对比度 ; When tS[k]≥D 亮度 When tS[k]<D 亮度 When , no operation is performed; When Db[k]≥D 对比度 When Db[k]<D 对比度 , no action is performed.

5. The method for identifying defects in LED circuit boards based on optical imaging according to claim 4, characterized in that: The process of extracting features from standard optical images and obtaining corresponding image optical feature information includes: Detect image key points in standard style optical images using SIFT algorithm; Using the Canny edge detection algorithm to identify the edge contours of the LED circuit board and the edge contours of the distributed circuits on the LED circuit board in the standard optical pattern, and combining them as the edge contour information of the standard optical pattern; Feature extraction of the standard optical pattern includes shape feature extraction, texture feature extraction, color feature extraction, and frequency domain feature extraction, thereby obtaining the geometric properties, texture information, color distribution information, and frequency domain characteristics of the LED circuit board and the circuits distributed on the LED circuit board corresponding to the current standard optical pattern; The image key points, edge contour information, geometric properties, texture information, color distribution information and frequency domain features corresponding to the standard style optical image are integrated and then used as the image optical feature information of the corresponding standard style optical image.

6. The method for identifying defects in LED circuit boards based on optical imaging according to claim 5, characterized in that: The process of building a defect recognition model includes: Construct the initial CNN model through convolutional neural network technology; Select LED circuit boards that have been classified into defect types and LED circuit boards without defects, obtain the corresponding optical feature information of their respective standard style optical images after processing, and build the information pooling layer and data output layer of the CNN model; The optical feature information is input into the information pooling layer to obtain the corresponding main optical features and the pooling coefficients corresponding to the main optical features. The part with the highest data proportion in the optical feature information is used as the main optical feature, and the defect judgment interval is set; Based on the subordinate relationship between the pooling coefficient of the main optical features and the defect judgment interval, it is determined whether the currently identified LED circuit board has defects and the type of defects identified. The recognition result is output by the data output layer, and the recognition result is compared with the actual result. Based on the comparison result, it is decided whether to train the CNN model, and then the final defect recognition model is constructed.

7. The method for identifying defects in LED circuit boards based on optical imaging according to claim 6, characterized in that: The process of inputting the image optical feature information into the defect recognition model and then classifying and identifying the corresponding defect type of the LED circuit board includes: When the defect recognition model is successfully constructed, the image optical feature information corresponding to the currently unclassified LED circuit board is input into the defect recognition model to classify and identify the defect types of the LED circuit board, including solder joint defects, electrical connection defects, circuit defects, and interlayer defects.

8. The method for identifying defects in LED circuit boards based on optical imaging according to claim 7, characterized in that: The defect recognition model's classification and identification results for LED circuit boards are fed back to the corresponding operation terminal of the production line. The operation terminal controls the robotic arm deployed on the production line to sort out abnormal LED circuit boards. The process includes: The defect recognition model feeds back the real-time recognition results to the corresponding operation terminal of the production line, which then locates the defective LED circuit board and the corresponding circuit board storage area on the production line. Each circuit board storage area is equipped with a corresponding robotic arm. When the identification result is received that the LED circuit board has a defect, the corresponding LED circuit board is marked as an abnormal LED circuit board. The robotic arm is controlled to pick out the abnormal LED circuit boards in each circuit board storage area and classify and store them according to the defect type.

Images