Panel zero sample defect detection method, system and equipment and storage medium
Through the deep learning network, the panel image features are extracted, the similarity is calculated and the abnormal map is constructed, which solves the problem of easy passing of defect detection in the panel manufacturing process, and achieves efficient and accurate zero-sample defect detection.
Patent Information
- Application Number
- CN202411426629.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-14
- Publication Date
- 2025-02-07
AI Technical Summary
The prior art is prone to over-checking in the defect detection of regular feature panel images during panel manufacturing, and requires a large number of manual labeling of samples, resulting in high costs.
The deep semantic features of the panel image are extracted through the deep learning network, the similarity between each feature point and the remaining feature point is calculated, the feature anomaly map is constructed, and statistics are performed to obtain the abnormal area and realize defect detection.
Defect detection can be carried out without model training, with high detection accuracy and high efficiency, avoiding manual annotation, reducing detection costs, and suitable for zero-sample defect detection.
Smart Images
Figure CN119399519A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of industrial defect detection, and in particular to a panel zero-sample defect detection method, system, device and storage medium. Background Art
[0002] In the panel manufacturing process, for panel images with regular features, such as Array, CF, OLED and other scenes, the existing technology mainly captures anomalies through sliding matching. This method requires confirmation of the matching periodic pixel values, and correction processing is required for situations such as image periodic tilt, which is prone to over-inspection. At the same time, the existing technology can also perform deep learning training in a supervised manner to learn defective areas from panel images. This method requires the annotation of a large number of learning samples, and the cost of manual annotation is relatively high. Summary of the invention
[0003] In order to solve the problem that panel images with existing regular features are prone to over-detection during defect detection, embodiments of the present invention provide a panel zero-sample defect detection method, system, device and storage medium.
[0004] In a first aspect, an embodiment of the present invention provides a panel zero-sample defect detection method, the method comprising the following steps:
[0005] Extract features of the panel image based on the deep learning network to obtain feature vectors of all feature points of the panel image;
[0006] Based on the feature vectors of all feature points in the panel image, similarity calculation is performed on all feature points to obtain similarity values of feature point pairs;
[0007] Anomaly calculation is performed on all feature points based on the similarity values of feature point pairs to construct a feature anomaly map about the feature points;
[0008] The panel image is segmented into abnormal regions based on the characteristic abnormality map of the feature points to obtain the defective regions of the panel image.
[0009] In the above embodiment, the present invention is based on the periodic characteristics of the panel production and manufacturing process. First, deep semantic features are extracted by using a neural network to calculate the similarity between each feature point and the remaining feature points. Then, the outlier value of the current feature point is determined to form a feature anomaly map. Finally, a statistical algorithm is used to perform statistics on the feature anomaly map to obtain the abnormal area, that is, to detect the defective area in the panel image. There is no need for model training, and defective area detection can be performed directly. The defect detection has high accuracy and efficiency, does not require manual labeling, and has low detection cost. It is suitable for defect detection with low incidence and zero sample.
[0010] As some optional implementations of the present application, the feature vectors of all feature points of the panel image are expressed as:
[0011] G={F ij |i=1,...,w g ,j=1,...,h g}
[0012] Among them, G represents the set of feature vectors of all feature points, F ij Represents the feature vector, i represents the width corresponding to the location of the feature point, and j represents the height corresponding to the location of the feature point.
[0013] In the above embodiment, the present invention automatically extracts hierarchical and abstract feature vectors from the original panel image through models such as convolutional neural networks, and these feature vectors form a set of all feature vectors of the current panel image through layer-by-layer transmission and optimization.
[0014] As some optional implementations of the present application, the feature vector includes image information of the area around the feature point; wherein the image information of the area around the feature point includes color, texture and shape.
[0015] In the above-mentioned embodiment, the present invention can quickly and accurately find abnormalities of feature points by analyzing image information of the area surrounding the feature points.
[0016] As some optional implementations of the present application, the similarity calculation method includes distance measurement and vector similarity method.
[0017] In the above embodiment, the present invention uses distance measurement and vector similarity measurement to perform similarity calculation, wherein the distance measurement method includes but is not limited to Euclidean distance, Manhattan distance, Minkowski distance, etc., and the most similar feature point pairs can be quickly obtained through similarity calculation.
[0018] As some optional implementations of the present application, the process of performing anomaly calculation on all feature points based on the similarity values of feature point pairs to construct a feature anomaly map about the feature points is as follows:
[0019] Perform similarity matching on the feature points according to the similarity values between each feature point and the remaining feature points to obtain the k feature point pairs with the highest similarity matching values;
[0020] Calculate the outlier value of each feature point according to the similarity values of the k feature point pairs with the highest similarity matching values, wherein the outlier value is the average of the similarity values of the k feature point pairs;
[0021] The image is constructed according to the outlier values of all feature points of the panel image to obtain the feature anomaly map of all feature points.
[0022] In the above embodiment, the present invention constructs a feature anomaly map based on the abnormal values of all feature points. The feature anomaly map can reflect the abnormal area of the panel image, thereby realizing defect detection.
[0023] As some optional implementations of the present application, the calculation formula of the outlier value is:
[0024]
[0025] Among them, s ij represents the outlier value of the feature point, k represents the number of feature point pairs, and F ij represents the feature vector, and f represents the feature vector of the feature point to the remaining feature points.
[0026] In the above embodiment, the outlier value of the feature point of the present invention is the average of the similarity values of the k most similar feature point pairs. The use of the average value can reduce the interference of noise on the feature point anomaly detection. At the same time, only the similarity values of the k most similar feature point pairs related to the feature points are selected to calculate the outlier value, which can simplify the parameters and improve the efficiency of anomaly detection.
[0027] As some optional implementations of the present application, the characteristic anomaly map is upsampled to the panel image size by linear interpolation, and Gaussian filtering is used to remove noise points.
[0028] In the above embodiment, the present invention upsamples the characteristic anomaly map to the size of the panel image, which can facilitate the positioning of the abnormal area, and prevents the interference of noise points on anomaly detection by Gaussian filtering.
[0029] As some optional implementations of the present application, the process of segmenting the panel image into abnormal regions based on the characteristic abnormality map of the characteristic points to obtain the defective regions of the panel image is as follows:
[0030] All abnormal feature points in the feature anomaly graph are detected using mean standard deviation statistics;
[0031] The defective area is marked in the panel image according to the position information of all abnormal feature points in the feature anomaly map.
[0032] In the above embodiment, the present invention uses the mean standard deviation statistics method to quickly obtain abnormal feature points, and then mark the defect area from the panel image.
[0033] In a second aspect, the present invention provides a panel zero-sample defect detection system, the system comprising:
[0034] A feature point extraction unit, wherein the feature point extraction unit performs feature extraction on the panel image based on a deep learning network to obtain feature vectors of all feature points of the panel image;
[0035] A similarity calculation unit, wherein the similarity calculation unit performs similarity calculation on all feature points based on feature vectors of all feature points of the panel image to obtain similarity values of feature point pairs;
[0036] an anomaly calculation unit, wherein the anomaly calculation unit performs anomaly calculation on all feature points based on similarity values of feature point pairs to construct a feature anomaly map about the feature points;
[0037] The abnormal region segmentation unit performs abnormal region segmentation on the panel image based on the characteristic abnormality map of the characteristic points to obtain the defect region of the panel image.
[0038] In a third aspect, the present invention provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the panel zero-sample defect detection method.
[0039] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, and the computer program, when executed by a processor, implements the panel zero-sample defect detection method.
[0040] The beneficial effects of the present invention are as follows: Based on the periodic characteristics in the panel production and manufacturing process, the present invention first extracts deep semantic features by using a neural network, calculates the similarity between each feature point and the remaining feature points, and then determines the outlier value of the current feature point to form a feature anomaly map. Finally, a statistical algorithm is used to perform statistics on the feature anomaly map to obtain the abnormal area, that is, to detect the defective area in the panel image. No model training is required, and defective area detection can be performed directly. The defect detection has high accuracy and efficiency, does not require manual labeling, and has low detection cost. It is suitable for defect detection with low incidence and zero sample. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.
[0042] Figure 1 is a schematic diagram of the computer device structure of the hardware operating environment described in the embodiment of the present invention;
[0043] Figure 2 is a flow chart of a panel zero-sample defect detection method according to an embodiment of the present invention;
[0044] Figure 3 is a schematic diagram of a panel image according to an embodiment of the present invention;
[0045] Figure 4 is a schematic diagram of a characteristic anomaly graph according to an embodiment of the present invention;
[0046] Figure 5 It is a structural block diagram of the panel zero-sample defect detection system described in an embodiment of the present invention. DETAILED DESCRIPTION
[0047] It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0048] In order to solve the problem that panel images with existing regular features are prone to over-detection during defect detection, the embodiment of the present invention provides a panel zero-sample defect detection method, system, device and storage medium. Before introducing the specific technical solution of the present application, the hardware operating environment involved in the embodiment of the present application is first introduced.
[0049] See also Figure 1 , Figure 1 The figure is a schematic diagram of the computer device structure of the hardware operating environment involved in the embodiment of the present invention.
[0050] like Figure 1 As shown, the computer device may include: a processor, such as a central processing unit (CPU), a communication bus, a user interface, a network interface, and a memory. Among them, the communication bus is used to realize the connection and communication between these components. The user interface may include a display screen (Display), an input unit such as a keyboard (Keyboard), and optionally, the user interface may also include a standard wired interface and a wireless interface. The network interface may be a wired interface or a wireless interface (such as a wireless fidelity interface). The memory may be a high-speed random access memory (Random Access Memory, RAM), or a stable non-volatile memory (Non-Volatile Memory, NVM), such as a disk storage. The memory may optionally be a storage device independent of the aforementioned processor.
[0051] Those skilled in the art will understand that Figure 1 The structure shown in the figure does not constitute a limitation on the computer device, and may include more or less components than shown in the figure, or combine some components, or arrange the components differently.
[0052] like Figure 1 As shown, the memory as a storage medium may include an operating system module, a network communication module, a user interface module and a data storage module.
[0053] exist Figure 1 In the computer device shown, the network interface is mainly used for data communication with the network server; the user interface is mainly used for data interaction with the user; the processor and memory in the computer device of the present application can be set in the computer device, and the computer device calls the panel zero sample defect detection system stored in the data storage module through the processor, and executes the panel zero sample defect detection method provided in the embodiment of the present application.
[0054] Based on the hardware environment of the aforementioned embodiment, the embodiment of the present application provides a panel zero-sample defect detection method, see Figure 2 , Figure 2 The flowchart of the panel zero-sample defect detection method is as follows:
[0055] (1) Input the panel image to be detected into the deep learning network, and then extract the features of the panel image based on the deep learning network to obtain the feature vectors of all feature points of the panel image, and form the feature set of the current panel image, G = {F ij |i=1,...,w g ,j=1,...,h g}; where G represents the set of feature vectors of all feature points, F ij represents a feature vector with a dimension of m, i represents the width corresponding to the position of the feature point, and j represents the height corresponding to the position of the feature point; wherein the deep neural network includes but is not limited to a convolutional neural network (CNN), a deep residual network (Res-Net), a feature pyramid network (FPN), etc.; preferably, the embodiment of the present invention uses a convolutional neural network to extract features from the panel image.
[0056] In the embodiment of the present invention, the feature points of the panel image are usually relatively significant or regular points in the panel image, such as corner points, edge points, and extreme value points.
[0057] In the embodiment of the present invention, a feature vector with a dimension of m is a vector containing m values or elements, which describe the measurement of the feature point on certain features or attributes. Specifically, the feature vector contains image information of the area around the feature point, wherein the image information of the area around the feature point includes but is not limited to color, texture, shape, etc.
[0058] (2) Calculate the similarity of all feature points based on the feature vectors of all feature points in the panel image to obtain the similarity values of feature point pairs.
[0059] In an embodiment of the present invention, the similarity calculation method includes a distance measurement and a vector similarity measurement method, wherein the distance measurement method includes but is not limited to Euclidean distance, Manhattan distance, Minkowski distance, etc.; the vector similarity measurement method includes but is not limited to cosine similarity, etc.; preferably, the embodiment of the present invention adopts the Euclidean distance method to calculate the similarity of all feature points; specifically, the feature point pair represents the combination of each feature point and each remaining feature point.
[0060] Specifically, the calculation formula of the Euclidean distance is as follows:
[0061]
[0062] Where D(X,Y) represents the Euclidean distance between feature vector X and feature vector Y, X = (x1, x2, ..., x m ) and Y=(y1,y2,...,y m ) represents two feature vectors of dimension m.
[0063] (3) Anomaly calculation is performed on all feature points according to the similarity values of feature point pairs to construct a feature anomaly map about the feature points.
[0064] In the embodiment of the present invention, the process of calculating all feature point anomalies is as follows:
[0065] (3.1) performing similarity matching on the feature points according to the similarity values between each feature point and the remaining feature points to obtain k feature point pairs with the highest similarity matching values; wherein the similarity matching is to quickly find the most similar k feature point pairs by performing a fast approximate nearest neighbor search.
[0066] (3.2) Calculate the outlier value of each feature point based on the similarity values of the k feature point pairs with the highest similarity matching values, where the outlier value is the average of the similarity values of the k feature point pairs.
[0067] (3.3) Image construction is performed based on the outlier values of all feature points of the panel image to obtain the feature anomaly map of all feature points, where the size of the feature anomaly map is w g *h g .
[0068] Specifically, the calculation formula of the outlier value is as follows:
[0069]
[0070] Among them, s ij represents the outlier value of the feature point, k represents the number of feature point pairs, and F ijrepresents the feature vector, and f represents the feature vector of the feature point to the remaining feature points.
[0071] In the embodiment of the present invention, the construction process of the characteristic anomaly map is as follows:
[0072] ① Initialize vertices: Each vertex corresponds to each feature point.
[0073] ② Construct undirected edges and set weights: For any two vertices in the feature anomaly graph, construct an undirected edge and set the weight of the directed edge, where the weight of the undirected edge is the similarity of the corresponding feature point pair.
[0074] ③Construct self-loop edges and set weights: For the vertices in the feature anomaly graph, set their weights to the anomaly values of the feature points.
[0075] (4) Segment the abnormal area of the panel image based on the characteristic abnormal map of the feature points to obtain the defect area of the panel image.
[0076] In the embodiment of the present invention, the process of segmenting the abnormal region is as follows:
[0077] (4.1) All abnormal feature points in the feature anomaly map are detected by using mean standard deviation statistics; specifically, a parameter value p is set, and when the pixel value of the corresponding feature point is greater than μ+pσ, the feature point is determined to be abnormal, wherein μ is the mean of the pixel values of all feature points, and σ is the standard deviation of the pixel values of all feature points; specifically, the parameter value p can be set according to actual conditions, and the embodiment of the present invention does not limit the specific numerical value of the parameter value P.
[0078] (4.2) Mark the defect area in the panel image according to the position information of all abnormal feature points in the feature anomaly map. The feature anomaly map is upsampled to the panel image size by linear interpolation, and the noise points are removed by Gaussian filtering. Figure 3 , Figure 3 For a schematic diagram of the panel image, see Figure 4 , a schematic diagram of the characteristic anomaly graph.
[0079] In addition, in order to further improve the accuracy of defect detection, the embodiment of the present invention can significantly improve the accuracy of defect detection by performing a secondary verification on the characteristic anomaly graph on the basis of the above-mentioned defect detection.
[0080] In the embodiment of the present invention, the process of performing a secondary check on the characteristic anomaly graph is as follows:
[0081] (4.3) Constructing a characteristic anomaly map of the normal panel image based on the normal panel image; wherein, the construction process of the characteristic anomaly map of the normal panel image can refer to the construction process of the characteristic anomaly map of the panel image to be detected.
[0082] (4.4) Subtract the pixel values of the feature anomaly map corresponding to the normal panel image and the feature anomaly map corresponding to the image to be detected to obtain a final anomaly MAP map; wherein the anomaly MAP map no longer contains the panel image background interference.
[0083] (4.5) The defective area is marked in the panel image based on the location information of all abnormal feature points in the final abnormal MAP.
[0084] In an embodiment of the present invention, by subtracting the pixel values of the characteristic abnormality map corresponding to the normal panel image and the characteristic abnormality map corresponding to the image to be detected, an abnormality MAP map without background interference can be obtained. The abnormality MAP map can be used to more accurately determine and locate abnormalities, and thus the defective area can be obtained from the panel image to be detected.
[0085] In summary, the embodiment of the present invention is based on the periodic characteristics of the panel production and manufacturing process. First, deep semantic features are extracted by using a neural network, and the similarity between each feature point and the remaining feature points is calculated. Then, the outlier value of the current feature point is determined to form a feature anomaly map. Finally, a statistical algorithm is used to perform statistics on the feature anomaly map to obtain the abnormal area, that is, to detect the defective area in the panel image. The embodiment of the present invention does not require model training and can directly perform defective area detection. The defect detection has high accuracy and efficiency, does not require manual labeling, and has low detection cost. It is suitable for defect detection with low incidence and zero sample.
[0086] In addition, in one embodiment, based on the same inventive concept as the above embodiment, the embodiment of the present invention provides a panel zero-sample defect detection system, see Figure 5 , Figure 5 : is a structural block diagram of the panel zero-sample defect detection system, the system corresponds one-to-one to the method of embodiment 1, and the system includes:
[0087] A feature point extraction unit, wherein the feature point extraction unit performs feature extraction on the panel image based on a deep learning network to obtain feature vectors of all feature points of the panel image;
[0088] A similarity calculation unit, wherein the similarity calculation unit performs similarity calculation on all feature points based on feature vectors of all feature points of the panel image to obtain similarity values of feature point pairs;
[0089] an anomaly calculation unit, wherein the anomaly calculation unit performs anomaly calculation on all feature points based on similarity values of feature point pairs to construct a feature anomaly map about the feature points;
[0090] The abnormal region segmentation unit performs abnormal region segmentation on the panel image based on the characteristic abnormality map of the characteristic points to obtain the defect region of the panel image.
[0091] It should be noted that each unit in the panel zero-sample defect detection system in this embodiment corresponds one-to-one to each step in the panel zero-sample defect detection method in the aforementioned embodiment. Therefore, the specific implementation method and technical effects achieved by this embodiment can refer to the implementation method of the panel zero-sample defect detection method, and will not be repeated here.
[0092] In addition, in one embodiment, the present application further provides a computer device, which includes a processor, a memory, and a computer program stored in the memory, and the computer program implements the method in the aforementioned embodiment when executed by the processor.
[0093] In addition, in one embodiment, the present application further provides a computer storage medium, on which a computer program is stored, and the computer program implements the method in the aforementioned embodiment when executed by a processor.
[0094] In some embodiments, the computer readable storage medium may be a memory such as FRAM, ROM, PROM, EPROM, EEPROM, flash memory, magnetic surface memory, optical disk, or CD-ROM; or various devices including one or any combination of the above memories. The computer may be various computing devices including intelligent terminals and servers.
[0095] In some embodiments, executable instructions may be in the form of a program, software, software module, script or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as a stand-alone program or as a module, component, subroutine or other unit suitable for use in a computing environment.
[0096] As an example, executable instructions may, but need not, correspond to a file in a file system, may be stored as part of a file that stores other programs or data, such as in one or more scripts in a HyperText Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files storing one or more modules, subroutines, or code portions).
[0097] By way of example, executable instructions may be deployed to be executed on one computing device, or on multiple computing devices located at one site, or on multiple computing devices distributed across multiple sites and interconnected by a communication network.
[0098] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.
[0099] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.
[0100] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory / random access memory, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a multimedia terminal device (which can be a mobile phone, a computer, a television receiver, or a network device, etc.) to execute the methods described in each embodiment of the present application.
[0101] The above are only preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.
Claims
1. A panel zero-sample defect detection method, characterized in that: The method comprises the following steps: Extract features of the panel image based on the deep learning network to obtain feature vectors of all feature points of the panel image; Based on the feature vectors of all feature points in the panel image, similarity calculation is performed on all feature points to obtain similarity values of feature point pairs; Anomaly calculation is performed on all feature points based on the similarity values of feature point pairs to construct a feature anomaly map about the feature points; The panel image is segmented into abnormal regions based on the characteristic abnormality map of the feature points to obtain the defective regions of the panel image.
2. A panel zero-sample defect detection method according to claim 1, characterized in that: The feature vectors of all feature points of the panel image are expressed as: G={F ij |i=1,...,w g ,j=1,...,h g } Among them, G represents the set of feature vectors of all feature points, F ij Represents the feature vector, i represents the width corresponding to the location of the feature point, and j represents the height corresponding to the location of the feature point.
3. The panel zero-sample defect detection method according to claim 1, characterized in that: The feature vector contains image information of the area around the feature point, wherein the image information of the area around the feature point includes color, texture and shape.
4. The panel zero-sample defect detection method according to claim 1, characterized in that: The similarity calculation method includes distance measurement and vector similarity method.
5. The panel zero-sample defect detection method according to claim 1, characterized in that: The process of calculating the anomalies of all feature points based on the similarity values of feature point pairs to construct a feature anomaly map about the feature points is as follows: Perform similarity matching on the feature points according to the similarity values between each feature point and the remaining feature points to obtain the k feature point pairs with the highest similarity matching values; Calculate the outlier value of each feature point according to the similarity values of the k feature point pairs with the highest similarity matching values, wherein the outlier value is the average of the similarity values of the k feature point pairs; The image is constructed according to the outlier values of all feature points of the panel image to obtain the feature anomaly map of all feature points.
6. A panel zero-sample defect detection method according to claim 5, characterized in that: The calculation formula of the outlier value is: Among them, s ij represents the outlier value of the feature point, k represents the number of feature point pairs, and F ij represents the feature vector, and f represents the feature vector of the feature point to the remaining feature points.
7. The panel zero-sample defect detection method according to claim 5, characterized in that: The characteristic anomaly map is upsampled to the panel image size by linear interpolation, and Gaussian filtering is used to remove noise points.
8. The panel zero-sample defect detection method according to claim 1, characterized in that: The process of segmenting the abnormal area of the panel image based on the characteristic abnormal map of the feature points to obtain the defect area of the panel image is as follows: All abnormal feature points in the feature anomaly graph are detected using mean standard deviation statistics; The defective area is marked in the panel image according to the position information of all abnormal feature points in the feature anomaly map.
9. A panel zero-sample defect detection system, characterized in that: The system comprises: A feature point extraction unit, wherein the feature point extraction unit performs feature extraction on the panel image based on a deep learning network to obtain feature vectors of all feature points of the panel image; A similarity calculation unit, wherein the similarity calculation unit performs similarity calculation on all feature points based on feature vectors of all feature points of the panel image to obtain similarity values of feature point pairs; an anomaly calculation unit, wherein the anomaly calculation unit performs anomaly calculation on all feature points based on similarity values of feature point pairs to construct a feature anomaly map about the feature points; The abnormal region segmentation unit performs abnormal region segmentation on the panel image based on the characteristic abnormality map of the characteristic points to obtain the defect region of the panel image.
10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the panel zero-sample defect detection method described in any one of claims 1-8 is implemented.
11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for zero-sample defect detection of a panel described in any one of claims 1 to 8 is implemented.
Citation Information
Cited By
Panel zero-shot defect detection method and system, device, and storage medium
WO2026081607A1