Panel mura defect detection method and system and medium

Abstract

The invention belongs to the technical field of defect detection, and discloses a panel mura defect detection method and system and a medium, and the method comprises the steps: obtaining a panel image, carrying out the preprocessing of the panel image, and obtaining a gray image; performing wavelet decomposition on the grayscale image based on wavelet transform to obtain an image of a wavelet domain; enhancing the wavelet domain image based on an adaptive threshold function to obtain an enhanced wavelet domain image; converting the enhanced image of the wavelet domain into a spatial domain to obtain a spatial domain image; and post-processing the spatial domain image to obtain a mura defect detection result. According to the method, the grayscale image is converted to the wavelet domain, and the wavelet domain image is enhanced by using the adaptive threshold function, so that noise interference can be effectively eliminated; converting the image from a wavelet domain to a spatial domain, and post-processing the image in the spatial domain to overcome the brightness influence of the image background; therefore, the detection method provided by the invention can effectively improve the accuracy and robustness of mura defect detection.

Description

Technical Field

[0001] The present invention belongs to the technical field of defect detection, and particularly relates to a method, system and medium for detecting panel mura defects. Background Art

[0002] Mura defects are visual defects formed due to uneven brightness or chromaticity in a display, commonly found in screens such as LCD and OLED. They mainly appear as patchy or linear color differences. The defect area has a low contrast with the surrounding area, and the boundary is blurred. Usually, it needs to be observed at an inclined angle, especially obvious under a solid color background; thus, it will reduce the display quality and affect the user experience (such as "yellow spots" or light leakage on the screen).

[0003] Currently, there are the following two types of methods for detecting Mura defects: Instrument detection: Using an imaging luminance meter (such as Konica Minolta ProMetric series) or a surface luminance meter (RVS) to quantify the brightness difference; however, instrument detection relies on human eye perception, and the subjectivity of defects is strong, resulting in large errors in the detection results; Algorithm analysis: Based on polynomial surface fitting technology to separate the background and the defect area, and identify low-contrast Mura; due to the wide variety of product types of display panels, the method based on background modeling has poor robustness and is difficult to handle products of different manufacturing processes and different types of display panels.

[0004] Therefore, the existing Mura defect detection methods at least still have poor robustness and the detection accuracy needs to be improved. Summary of the Invention

[0005] The purpose of the present invention is to provide a method, system and medium for detecting panel mura defects, so as to solve the problem that the existing Mura defect detection methods at least still have poor robustness and the detection accuracy needs to be improved.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions: In the first aspect, the present invention provides a method for detecting panel mura defects, and the method includes: Obtain a panel image, preprocess the panel image to obtain a grayscale image; Perform wavelet decomposition on the grayscale image based on wavelet transform to obtain an image in the wavelet domain; Enhance the image in the wavelet domain based on an adaptive threshold function to obtain an enhanced image in the wavelet domain; Convert the enhanced image in the wavelet domain to the spatial domain to obtain a spatial domain image; Perform post-processing on the spatial domain image to obtain the mura defect detection result.

[0007] Preferably, the preprocessing at least includes: grayscale conversion processing and median filtering processing.

[0008] Preferably, enhancing the wavelet domain image based on an adaptive threshold function to obtain an enhanced image in the wavelet domain, including: Extracting the detail coefficients of the image in the wavelet domain; Enhancing the detail coefficients based on the adaptive threshold function to obtain enhanced detail coefficients; Based on the enhanced detail coefficients, obtaining the enhanced image in the wavelet domain.

[0009] Preferably, the expression of the adaptive threshold function is: ; ; In the formula, is the adaptive threshold function, is the threshold, N is the length of the wavelet coefficients of the wavelet transform, is the root mean square error of the wavelet coefficients of the wavelet transform, j is the number of layers of wavelet decomposition of the wavelet transform, is the k-th detail coefficient of the j-th layer, is the adjustable coefficient, ln() is the natural logarithm function, log 10 () is the logarithm function with base 10.

[0010] Preferably, the calculation expression of the enhanced detail coefficients is: ; In the formula, is the k-th detail coefficient of the j-th layer, is the enhanced detail coefficient, is the preset coefficient, sgn() is the sign function, is the adaptive threshold function, is the threshold.

[0011] Preferably, post-processing the spatial domain image to obtain the mura defect detection result, including: Adjusting the grayscale value of the spatial domain image to obtain an adjusted spatial domain image; Performing binarization processing on the adjusted spatial domain image to obtain a binary image; Extracting the white-value mura defects and black-value mura defects in the binary image; Based on the white-value mura defects and black-value mura defects, determining the mura defect characteristics and their positions, and using the mura defect characteristics and their positions as the mura defect detection result.

[0012] Preferably, the method further includes: Construct an evaluation function; Based on the evaluation function, evaluate the mura defect features to obtain the evaluation values of the mura defect features; Judge whether the evaluation value of the mura defect feature is less than the preset value. If so, eliminate the mura defect feature.

[0013] Preferably, the expression of the evaluation function is: ; In the formula, is the evaluation value of the i-th mura defect feature, is the area of the i-th mura defect feature, is the brightness of the i-th mura defect feature, is the background brightness.

[0014] In a second aspect, the present invention provides a panel mura defect detection system for implementing the above panel mura defect detection method. The system includes: An image acquisition module for acquiring a panel image, preprocessing the panel image to obtain a grayscale image; A first conversion module for performing wavelet decomposition on the grayscale image based on wavelet transform to obtain an image in the wavelet domain; An image enhancement module for enhancing the wavelet domain image based on an adaptive threshold function to obtain an enhanced image in the wavelet domain; A second conversion module for converting the enhanced image in the wavelet domain to the spatial domain to obtain a spatial domain image; An image processing module for post-processing the spatial domain image to obtain the mura defect detection result.

[0015] In a third aspect, the present invention provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the above panel mura defect detection method is implemented.

[0016] In a fourth aspect, the present invention provides a computer-readable storage medium with a computer program stored thereon. When the program is executed by a processor, the above panel mura defect detection method is implemented.

[0017] Beneficial effects: The present invention uses wavelet transform to transfer a grayscale image to the wavelet domain, and then uses an adaptive threshold function to enhance the image in the wavelet domain, which can effectively eliminate noise interference. Then, the image is transformed from the wavelet domain to the spatial domain, and post-processing is performed on the spatial domain image to overcome the influence of the brightness of the image background. Therefore, the detection method of the present invention can effectively improve the accuracy and robustness of mura defect detection. Description of the Drawings

[0018] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention, and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the embodiments of the present invention, but do not constitute a limitation to the embodiments of the present invention. In the drawings: Figure 1 is a flowchart of a method for detecting panel mura defects provided by an embodiment of the present invention; Figure 2 is a block diagram of a system for detecting panel mura defects provided by an embodiment of the present invention. Detailed Embodiments

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the present invention in combination with the drawings and the descriptions of the embodiments or the prior art. Obviously, the following descriptions of the structures of the drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings. It should be noted here that the descriptions of these embodiments are used to help understand the present invention, but do not constitute a limitation to the present invention.

[0020] Embodiment 1 Figure 1 is a flowchart of a method for detecting panel mura defects provided by an embodiment of the present invention. As Figure 1 shown, this embodiment provides a method for detecting panel mura defects, which runs on an intelligent detection device. The intelligent detection device can be a PC computer. The method includes: Step S10: Obtain a panel image, perform preprocessing on the panel image, and obtain a grayscale image.

[0021] In this embodiment, a CCD camera can be used to collect the image data of the display panel, and the image data of the display panel is uploaded to the intelligent detection device and stored in the memory of the intelligent detection device. When it is necessary to perform defect detection on the image data of the corresponding display panel, the image data of the corresponding display panel is extracted from the memory as the panel image to obtain the panel image. In this embodiment, the intelligent detection device can also perform real-time detection on the display panel. For example, the CCD camera is used to collect the image data of the display panel in real time. After the image data of the display panel is uploaded to the intelligent detection device, the intelligent detection device directly performs defect detection on the image data of the display panel to achieve real-time detection of panel mura defects.

[0022] In this embodiment, the preprocessing includes but is not limited to: gray-scale conversion processing and median filtering processing.

[0023] Among them, gray-scale conversion (Gray-level Transformation) is a basic technology of digital image processing. By converting a color image into a single-channel gray-scale image, effective extraction of brightness information and subsequent processing optimization are realized.

[0024] Among them, median filtering (Median Filtering) is a non-linear signal processing technology based on sorting statistics theory. The core algorithm logic is to replace the gray value of the target pixel of the image with the median of the gray values of all pixels within its neighborhood window. Compared with linear filtering (such as mean filtering), its non-linear characteristics enable it to achieve a better balance between noise suppression and edge preservation.

[0025] Step S20: Perform wavelet decomposition on the gray-scale image based on wavelet transform to obtain an image in the wavelet domain.

[0026] In this embodiment, wavelet transform (wavelet transform, WT) is a new transform analysis method. It inherits and develops the idea of localizing the short-time Fourier transform, and at the same time overcomes the disadvantages such as the window size not changing with frequency. It can provide a "time-frequency" window that changes with frequency and is an ideal tool for time-frequency analysis and processing of signals. Its main feature is that through the transform, certain aspects of the problem can be fully highlighted, and it can perform local analysis of time (space) frequency. Through stretching and translation operations, the signal (function) is gradually refined at multiple scales. Finally, time is finely divided at high frequencies and frequency is finely divided at low frequencies, and it can automatically adapt to the requirements of time-frequency signal analysis, so as to focus on any detail of the signal.

[0027] After the grayscale image is decomposed by wavelet transform, wavelet coefficients are obtained. The wavelet coefficients reflect the energy distribution characteristics of the grayscale image in multiple scales (frequency domain) and multiple positions (time / space domain). Among them, the scale dimension corresponds to different frequency components. Low-scale (high-frequency) coefficients (detail coefficients) reflect signal mutations or details, and high-scale (low-frequency) coefficients (approximate coefficients) characterize signal trends or overall characteristics.

[0028] Step S30: Enhance the wavelet-domain image based on the adaptive threshold function to obtain the enhanced image in the wavelet domain.

[0029] In this embodiment, since there is noise in wavelet transform, it is necessary to denoise and enhance the wavelet-domain image. The noise signal is generally a random signal, so its variance is often unknown. During the denoising process, it is necessary to estimate the threshold, select samples through a suitable method, and then estimate and select a threshold, and then retain the coefficients exceeding this threshold. The following are several common threshold estimation methods: fixed threshold, adaptive threshold based on the principle of unbiased likelihood estimation, and heuristic threshold.

[0030] In this embodiment, the core of wavelet-domain image denoising is the setting of the threshold and the threshold function. The threshold and the threshold function directly affect the quality and accuracy of wavelet-domain image denoising. For example, if the threshold is set too high, useful signals will be filtered out as noise; if it is set too low, the noise will not be filtered out thoroughly. Therefore, this embodiment realizes the denoising of the wavelet-domain image through the following steps: Step S301: Extract the detail coefficients of the image in the wavelet domain. In this implementation, the low-scale coefficients are extracted from the multi-scale (frequency domain) of the wavelet coefficients in step S20, and the low-scale coefficients are the detail coefficients.

[0031] Step S302: Enhance the detail coefficients based on the adaptive threshold function to obtain the enhanced detail coefficients.

[0032] In this embodiment, the expression of the adaptive threshold function is: ; ; In the formula, is the adaptive threshold function, is the threshold, N is the length of the wavelet coefficients of the wavelet transform, is the root mean square error of the wavelet coefficients of the wavelet transform, j is the number of layers of wavelet decomposition of the wavelet transform, is the k-th detail coefficient of the j-th layer, is the adjustable coefficient, ln() is the natural logarithm function, log 10 () is the logarithm function with base 10; among them, the value range of the adjustable coefficient is: ∈(0, 1).

[0033] In this embodiment, through the above threshold , more noise can be filtered out, improving the accuracy of absence detection.

[0034] As a further optimization of this embodiment, the calculation expression of the enhanced detail coefficient is: ; In the formula, is the k-th detail coefficient of the j-th layer, is the enhanced detail coefficient, is a preset coefficient, sgn() is the sign function, is the adaptive threshold function, is the threshold. Among them, the preset coefficient has a value range of: ∈[0, 1].

[0035] In this embodiment, the calculation function of the enhanced detail coefficient can achieve continuity at the threshold without generating discontinuities.

[0036] Moreover, the calculation function of the enhanced detail coefficient does not directly set the detail coefficient to zero, so that more useful detail coefficients can be retained, thereby improving the subsequent defect detection accuracy.

[0037] Secondly, the adjustable coefficient can also be adjusted, so that the value of F is adjustable, the value of the detail coefficient can be made larger, and at the same time, the preset coefficient can be adjusted, enabling the algorithm to adapt to the denoising of different products and improving the robustness of the algorithm.

[0038] Step S303: Based on the enhanced detail coefficient, obtain the enhanced image in the wavelet domain; In this embodiment, the wavelet domain image corresponding to the enhanced detail coefficient is used as the enhanced image in the wavelet domain.

[0039] Step S40: Convert the enhanced image in the wavelet domain to the spatial domain to obtain the spatial domain image; In this embodiment, the inverse wavelet transform can be used to convert the wavelet domain to the spatial domain to reconstruct the image in the spatial domain.

[0040] Step S50: Post-process the spatial domain image to obtain the mura defect detection result.

[0041] In this embodiment, post-processing the spatial domain image to obtain the mura defect detection result includes: Step S501: Adjust the gray values of the spatial domain image to obtain the adjusted spatial domain image; at this time, the adjusted spatial domain image contains mura defects, where the gray values of the spatial domain image include two types: white mura defects and black mura defects.

[0042] In this embodiment, the functional expression of the adjusted spatial domain image is: ; In the formula, is the gray value of the adjusted spatial domain image, is the coordinate of the gray value, is the gray value of the white mura defect, is the gray value of the black mura defect.

[0043] Among them, ; ; In the formula, is the gray value of the spatial domain image at the coordinate , is the background gray offset corresponding to the white mura defect, usually set to the median value of the image gray range, such as 128, is the background gray offset corresponding to the black mura defect, usually set to the maximum value of the image gray range, such as 255, and n is the number of bits of the spatial domain image.

[0044] Step S502: Perform binarization processing on the adjusted spatial domain image to obtain a binary image, separating the mura defects from the background.

[0045] Step S503: Extract the white-value mura defects and black-value mura defects in the binary image. The white-value mura defects are the white mura defects, and the black-value mura defects are the black mura defects.

[0046] Step S504: Based on the white-value mura defects and black-value mura defects, determine the mura defect features and their positions, and use the mura defect features and their positions as the mura defect detection results.

[0047] In this embodiment, morphological processing, such as disconnection processing, is performed on the binary image to reduce the influence of background noise, and at the same time, the positions of the mura defect features can be determined; among them, the size of the radius structural element for disconnection processing needs to be adjusted according to the size of the mura defects.

[0048] As a further optimization of this embodiment, the method further includes: Step a10: Construct an evaluation function; among them, the evaluation function is mainly constructed by parameters such as the area of mura defects, the brightness of defects, and the background brightness.

[0049] Step a20: Based on the evaluation function, evaluate the mura defect features to obtain the evaluation value of the mura defect features.

[0050] Step a30: Determine whether the evaluation value of the mura defect features is less than a preset value. If so, eliminate the mura defect features.

[0051] Therefore, the expression of the evaluation function is: ; In the formula, is the evaluation value of the i-th mura defect feature, is the area of the i-th mura defect feature, is the brightness of the i-th mura defect feature, is the background brightness.

[0052] In this embodiment, when the evaluation value of any mura defect feature is less than the preset value, it indicates that the severity of the mura defect is relatively small. And when the severity of the mura defect is relatively small, it cannot be observed by the human eye at this time. Such products can be regarded as medium and low-grade products; while for mura defect features that exceed the preset value, it indicates that the severity of the mura defect is relatively serious, and such products are defined as unqualified products; products without mura defect features are defined as high-grade qualified products to achieve the classification processing of products.

[0053] The present invention uses wavelet transform to transfer the grayscale image to the wavelet domain, and then uses an adaptive threshold function to enhance the wavelet domain image, which can effectively eliminate noise interference; then the image is converted from the wavelet domain to the spatial domain, and post-processing is performed on the spatial domain image to overcome the influence of the brightness of the image background; therefore, the detection method of the present invention can effectively improve the accuracy and robustness of mura defect detection.

[0054] Embodiment 2 Figure 2 is the block diagram of a panel mura defect detection system provided by an embodiment of the present invention. As Figure 2 shown, this embodiment provides a panel mura defect detection system for implementing the panel mura defect detection method in Embodiment 1. The system includes: An image acquisition module, used to acquire a panel image, preprocess the panel image, and obtain a grayscale image; A first conversion module, used to perform wavelet decomposition on the grayscale image based on wavelet transform to obtain an image in the wavelet domain; An image enhancement module, configured to enhance a wavelet domain image based on an adaptive threshold function to obtain an enhanced image in the wavelet domain; A second conversion module, configured to convert the enhanced image in the wavelet domain to the spatial domain to obtain a spatial domain image; An image processing module, configured to perform post-processing on the spatial domain image to obtain a mura defect detection result.

[0055] This embodiment also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the panel mura defect detection method in the first embodiment is implemented.

[0056] This embodiment also provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the panel mura defect detection method in the first embodiment is implemented.

[0057] The present invention uses wavelet transform to convert a grayscale image to the wavelet domain, and then uses an adaptive threshold function to enhance the wavelet domain image, which can effectively eliminate noise interference; then the image is converted from the wavelet domain to the spatial domain, and post-processing is performed on the spatial domain image to overcome the influence of the brightness of the image background; therefore, the detection method of the present invention can effectively improve the accuracy and robustness of mura defect detection.

[0058] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0059] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a system for implementing the functions specified in one Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0060] The above are only the embodiments of the present application and are not intended to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.

Claims

1. A method for detecting panel mura defects, characterized in that, The method comprises: Acquire a panel image, and preprocess the panel image to obtain a grayscale image; Based on wavelet transform, the grayscale image is decomposed into wavelet to obtain the image in wavelet domain; The wavelet domain image is enhanced based on the adaptive threshold function to obtain an enhanced image in the wavelet domain; Convert the enhanced image in the wavelet domain to the spatial domain to obtain a spatial domain image; The spatial domain image is post-processed to obtain the mura defect detection result.

2. The panel mura defect detection method according to claim 1, characterized in that The preprocessing at least includes: grayscale conversion processing and median filtering processing.

3. The panel mura defect detection method according to claim 1, wherein The wavelet domain image is enhanced based on the adaptive threshold function to obtain an enhanced image in the wavelet domain, including: Extract detail coefficients of images in wavelet domain; The detail coefficient is enhanced based on an adaptive threshold function to obtain an enhanced detail coefficient; Based on the enhanced detail coefficients, an enhanced image in the wavelet domain is obtained.

4. The panel mura defect detection method according to claim 3, wherein The expression of the adaptive threshold function is: ； ； In the formula, is the adaptive threshold function, is the threshold, N is the length of the wavelet coefficients of the wavelet transform, is the root mean square error of the wavelet coefficients of the wavelet transform, j is the number of layers of the wavelet decomposition of the wavelet transform, is the k-th detail coefficient of the j-th layer, is the adjustable coefficient, ln() is the natural logarithm function, log 10 () is the logarithm function with base 10.

5. The panel mura defect detection method according to claim 4, wherein The calculation expression of the enhanced detail coefficient is: ； wherein, is the k-th detail coefficient of the j-th layer, is the enhanced detail coefficient, is the preset coefficient, sgn() is the sign function, is the adaptive threshold function, is the threshold value.

6. The panel mura defect detection method according to claim 1, wherein, Post-process the spatial domain image to obtain the mura defect detection results, including: Adjusting the grayscale value of the spatial domain image to obtain an adjusted spatial domain image; Binarizing the adjusted spatial domain image to obtain a binary image; Extract white value mura defects and black value mura defects in binary images; Based on the white value mura defect and the black value mura defect, the mura defect characteristics and the position are determined, and the mura defect characteristics and the position are used as the mura defect detection result.

7. The panel mura defect detection method according to claim 1, characterized in that The method further comprises: Construct an evaluation function; Based on the evaluation function, the mura defect feature is evaluated to obtain an evaluation value of the mura defect feature; It is determined whether the evaluation value of the mura defect feature is less than a preset value. If so, the mura defect feature is removed.

8. The panel mura defect detection method according to claim 7, wherein The expression of the evaluation function is: ； Wherein, is the evaluation value of the i-th mura defect feature, is the area of the i-th mura defect feature, is the brightness of the i-th mura defect feature, is the background brightness.

9. A panel mura defect detection system for implementing the panel mura defect detection method according to any one of claims 1-8, characterized in that, The system comprises: An image acquisition module is used to acquire a panel image and preprocess the panel image to obtain a grayscale image; The first conversion module is used to perform wavelet decomposition on the grayscale image based on wavelet transform to obtain an image in wavelet domain; An image enhancement module is used to enhance the wavelet domain image based on an adaptive threshold function to obtain an enhanced image in the wavelet domain; The second conversion module is used to convert the enhanced image in the wavelet domain into the spatial domain to obtain a spatial domain image; The image processing module is used to post-process the spatial domain image to obtain the mura defect detection result.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by a processor, the panel mura defect detection method described in any one of claims 1 to 8 is implemented.

Images