Method and system for detecting non-uniform brightness defect of liquid crystal panel

Abstract

The invention relates to the technical field of liquid crystal panel defect detection, and particularly discloses a liquid crystal panel brightness non-uniform defect detection method and system, an image of a liquid crystal panel in a display mode is obtained by using a photometer, brightness information is captured through a high-precision sensor, and the brightness non-uniform defect of the liquid crystal panel is obtained by using an image segmentation algorithm. In combination with adaptive threshold segmentation, deep learning semantic segmentation and a region growing algorithm, regions with uneven brightness problems are accurately identified and extracted, brightness difference degrees, spatial distribution and time stability are analyzed respectively, corresponding feature indexes are calculated, the feature indexes are integrated, and the brightness difference degree, the spatial distribution and the time stability are calculated. The comprehensive score of each region with uneven brightness is generated through standardized calculation, the regions are divided into slight, medium and serious defects, the method is combined with real-time monitoring, the detection time and parameter setting can be dynamically adjusted, intelligent detection is carried out for different types of uneven brightness defects, the detection efficiency and precision are optimized, and redundant detection is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of liquid crystal panel defect detection, and in particular to a method and system for detecting uneven brightness defects of a liquid crystal panel. Background Art

[0002] Liquid crystal panels are widely used in televisions, monitors, smart phones and other devices, and their display effects directly affect the user experience. In the production process of liquid crystal panels, uneven brightness is one of the common defects. Uneven brightness defects may be caused by quality problems of liquid crystal materials, unstable panel manufacturing processes, improper control circuit design, etc. This defect not only affects the display effect, but also may shorten the service life of the liquid crystal panel. In order to ensure the display quality and consistency of liquid crystal panels, accurate uneven brightness defect detection is crucial. Therefore, how to effectively identify and classify different types of uneven brightness defects has become an important topic in the field of liquid crystal panel production and quality control.

[0003] The prior art has the following deficiencies: At present, most of the detection methods for uneven brightness defects of LCD panels are based on threshold judgment of static brightness differences, lacking a comprehensive analysis of the multi-dimensional characteristics of uneven brightness areas. These methods only focus on brightness differences, ignoring the spatial distribution of defects, the temporal stability of brightness changes, and the complexity of defect areas. In addition, existing detection systems usually use fixed detection time and parameter settings, and cannot dynamically adjust the detection strategy according to different types of defects. Therefore, when encountering complex brightness unevenness problems, traditional methods cannot effectively respond, resulting in insufficient detection accuracy, and misjudgment and missed judgment are relatively common. Moreover, in the case of large-scale production and high-frequency detection, fixed detection parameters will lead to redundant detection processes, increasing detection time and reducing efficiency. These deficiencies have seriously affected the accuracy and production efficiency of LCD panel quality control, and there is an urgent need for an intelligent detection method that can dynamically adapt to different types of uneven brightness defects. Summary of the invention

[0004] The object of the present invention is to provide a method and system for detecting uneven brightness defects of a liquid crystal panel to solve the above-mentioned problems.

[0005] The purpose of the present invention can be achieved through the following technical solutions: A method for detecting uneven brightness defects of a liquid crystal panel comprises the following steps: S1: using a photometer to obtain an image of the liquid crystal panel in a display mode, and preprocessing the image to reduce noise and optimize contrast; S2: Identify and separate the area containing uneven brightness problem through image segmentation algorithm, analyze the brightness difference degree, spatial distribution and temporal stability of the area containing uneven brightness problem respectively, and calculate the spatial distribution characteristic index, brightness difference characteristic index and temporal stability characteristic index of the area containing uneven brightness problem according to the analysis results; S3: Comprehensively analyze the spatial distribution characteristic index, brightness difference characteristic index and temporal stability characteristic index of the extracted brightness unevenness problem area, and generate a comprehensive score for each brightness unevenness area based on the analysis results; S4: Each uneven brightness area is classified into slight defect, medium defect and severe defect according to the comprehensive score of each uneven brightness area; S5: Combined with real-time monitoring, the detection time and parameter settings of different types of defects are dynamically adjusted to achieve intelligent detection of different types of uneven brightness defects.

[0006] As a further solution of the present invention: the method of identifying and separating the area containing the problem of uneven brightness by using the image segmentation algorithm specifically includes: Through the adaptive threshold segmentation method, the brightness threshold is determined based on the brightness distribution of the image, and the area with brightness difference is extracted from the image; The image is processed based on a semantic segmentation network of deep learning, and a convolutional neural network is used to identify areas with uneven brightness. The convolutional neural network is trained by annotated data sets to learn the brightness characteristics of the liquid crystal panel for area recognition; The region growing algorithm is used to dynamically expand the region boundary according to the brightness difference of adjacent pixels and extract the problem area of ​​uneven brightness.

[0007] As a further solution of the present invention: the analysis of the brightness difference degree, spatial distribution and temporal stability of the brightness unevenness problem area specifically includes: Analyze the degree of brightness difference in the area with uneven brightness problem, analyze the degree of brightness difference in the area with uneven brightness problem, and calculate the brightness difference characteristic index according to the amplitude of brightness fluctuation in the area with uneven brightness problem; Analyze the spatial distribution of the brightness unevenness problem area and extract the geometric characteristics of the spatial distribution, including the area, perimeter, length and width of the brightness unevenness problem area. According to the abnormality of the spatial distribution geometric characteristics, calculate the spatial distribution characteristic index; The temporal stability of the area with uneven brightness problem is analyzed, the stability of brightness change in the area with uneven brightness problem is analyzed, and the temporal stability characteristic index is calculated based on the analysis results, so as to evaluate the temporal stability of the area with uneven brightness problem.

[0008] As a further solution of the present invention: the process of obtaining the brightness difference characteristic index is: Apply discrete wavelet transform to the preprocessed image to decompose the image into low-frequency and high-frequency parts, including: Perform multi-level wavelet decomposition on the image to obtain the low-frequency component and high-frequency component of each layer; The calculation expression of each layer of wavelet decomposition is: ; ; in, is the low frequency component, is the high frequency component, and For low-pass and high-pass filters, represents the original image, Represents the pixel coordinates in the image; Calculate the brightness difference of each high-frequency component. The calculation expression is: ; in, Represents the pixel value of the high-frequency component, represents the mean value of high frequency components, represents the pixels in the high-frequency component, Represents the total number of pixels in the high-frequency component; Based on the calculated brightness difference, the brightness difference characteristic index of each high-frequency component is calculated, and the calculation expression is: ; in, is the total number of high-frequency subbands analyzed, represents the high frequency subband, Indicates The brightness difference of the high-frequency sub-bands, Indicates the brightness difference characteristic index of high-frequency components; The brightness difference characteristic index of the wavelet decomposition results of all levels is integrated to obtain the final brightness difference characteristic index. The calculation expression is: ; In the formula, represents the total number of wavelet decomposition layers, represents the wavelet decomposition layer, Represents the brightness difference characteristic index.

[0009] As a further solution of the present invention: the process of obtaining the spatial distribution characteristic index is: The area with uneven brightness is extracted, and each pixel is regarded as a node in the graph. Adjacent pixels are connected by edges, and the weight of the edge is represented by the difference in pixel brightness, including: Each edge in the graph connects two adjacent pixels, and the weight of the edge is Represents pixels and pixels The brightness difference between them is calculated as: ; in, Representation Node The brightness value of the corresponding pixel, Representation Node The brightness value of the corresponding pixel; Extract the geometric features of the area with uneven brightness, including: The area of ​​the region is calculated by the number of nodes in the uneven brightness area in the graph. The calculation expression is: ; in, is the number of nodes in the graph, Indicates the area of ​​uneven brightness; Calculate the perimeter of the area with uneven brightness. The calculation expression is: ; in, Indicates the perimeter of the area with uneven brightness, represents the edge set in the graph, Represents the indicator function. If the two nodes connected by the edge belong to the inside and outside of the region respectively, ,otherwise ; Calculate the aspect ratio of the area with uneven brightness. The calculation expression is: ; in, Represents the length of the circumscribed rectangle, Indicates the width of the bounding rectangle. Indicates the aspect ratio of the area with uneven brightness; According to the geometric characteristics of the uneven brightness area, the spatial distribution characteristic index is calculated comprehensively, and the calculation expression is: ; In the formula, Represents the spatial distribution characteristic index.

[0010] As a further solution of the present invention: the process of obtaining the temporal stability characteristic index is: Get the brightness values ​​of the area with uneven brightness at multiple time points to form a time series: ; in, Indicates the collection time point, Indicates the length of the time series; Through Fourier transformation, the time series Convert to frequency domain signal , the calculation expression is: ; in, represents the frequency domain signal, represents the imaginary unit, represents the logarithm of the base of a natural number, represents the frequency component; Calculate the amplitude spectrum of the frequency domain signal. The calculation expression is: ; in, Represents the frequency components in the frequency domain The corresponding amplitude; Find the main frequency and frequency energy of brightness changes by analyzing the amplitude spectrum; Among them, the main frequency is the frequency component corresponding to the maximum amplitude in the amplitude spectrum, which is recorded as ; Calculate the sum of the energies of all frequency components in the amplitude spectrum to obtain the total frequency energy. The calculation expression is: ; in, represents the total frequency energy, Indicates the maximum frequency of the frequency component; According to the calculation of total frequency energy and main frequency, the time stability characteristic index is calculated, and the calculation expression is: ; in, represents the temporal stability characteristic index, Representing time series The maximum amplitude, Indicates the main frequency.

[0011] As a further solution of the present invention: the spatial distribution characteristic index, brightness difference characteristic index and time stability characteristic index of the extracted brightness uneven problem area are comprehensively analyzed, and a comprehensive score is generated for each brightness uneven area according to the analysis results, specifically including: Obtain the spatial distribution characteristic index, brightness difference characteristic index and time stability characteristic index of the brightness unevenness problem area, perform normalized calculation processing on the spatial distribution characteristic index, brightness difference characteristic index and time stability characteristic index, and calculate the comprehensive score of the brightness unevenness area through the comprehensive score calculation expression; The comprehensive score calculation expression is: ; in, Represents the standardized value of the spatial distribution characteristic index, represents the normalized value of the brightness difference characteristic index, represents the normalized value of the temporal stability characteristic index, Indicates the overall rating.

[0012] As a further solution of the present invention: the above-mentioned classification of each uneven brightness area into slight defects, medium defects and severe defects according to the comprehensive score of each uneven brightness area specifically includes: According to the comprehensive score calculation expression, the comprehensive score of each uneven brightness area is calculated respectively, and the comprehensive score of each uneven brightness area is compared with the first threshold value. If the comprehensive score is greater than or equal to the first threshold value, it means that the defect state of the corresponding area is slight, which is recorded as a slight defect; If the comprehensive score is less than the first threshold, the comprehensive score is compared with the second threshold. If the comprehensive score is less than or equal to the second threshold, it means that the defect state of the corresponding area is serious and is recorded as a serious defect; If the comprehensive score is greater than the second threshold and less than the first threshold, it means that the defect state of the corresponding area is medium, which is recorded as medium defect; The value of the first threshold is greater than the value of the second threshold.

[0013] As a further solution of the present invention: the above-mentioned real-time monitoring is combined to dynamically adjust the detection time and parameter settings of different types of defects to realize intelligent detection of different types of uneven brightness defects, specifically including: By monitoring the changes in the uneven brightness area of ​​the LCD panel in real time, the defect types are classified according to the brightness difference characteristics, spatial distribution characteristics and time stability characteristics; When the system detects that the uneven brightness defect of the LCD panel in the area is a serious defect, it will automatically extend the detection time and increase the detection accuracy; if the uneven brightness defect of the LCD panel is a minor defect, the system will automatically shorten the detection time and reduce redundant detection. When the system detects that the uneven brightness defect of the LCD panel in the area is a medium defect, it will be detected according to the original detection time.

[0014] A system for detecting uneven brightness defects of a liquid crystal panel, comprising: A data acquisition module, which uses a photometer to acquire an image of the liquid crystal panel in a display mode and preprocesses the image to reduce noise and optimize contrast; A region recognition module, which uses an image segmentation algorithm to identify and separate regions containing uneven brightness problems; A regional feature analysis module, which analyzes the brightness difference degree, spatial distribution and temporal stability of the brightness uneven problem area, and calculates the spatial distribution feature index, brightness difference feature index and temporal stability feature index of the brightness uneven problem area according to the analysis results; A comprehensive analysis module, which performs a comprehensive analysis on the spatial distribution characteristic index, the brightness difference characteristic index and the temporal stability characteristic index of the extracted brightness unevenness problem area, and generates a comprehensive score for each brightness unevenness area according to the analysis results; A defect recognition and classification module, wherein the defect recognition and classification module classifies each uneven brightness area into slight defects, medium defects and severe defects according to the comprehensive score of each uneven brightness area; A dynamic detection module combines real-time monitoring to dynamically adjust the detection time and parameter settings of different types of defects to achieve intelligent detection of different types of uneven brightness defects.

[0015] Beneficial effects of the present invention: (1) The present invention achieves accurate identification, classification and quantitative evaluation of the brightness unevenness problem area by comprehensively analyzing the multi-dimensional characteristics of the liquid crystal panel brightness unevenness defect (including spatial distribution characteristic index, brightness difference characteristic index and time stability characteristic index). Unlike the traditional detection method that relies only on a single brightness difference index, the present invention provides a comprehensive and dynamic analysis framework by combining multiple indicators, such as the geometric characteristics of spatial distribution (area, perimeter, aspect ratio) and the frequency characteristics of brightness fluctuation (such as brightness difference amplitude and time stability). This multi-dimensional evaluation method can not only capture the characteristics of various types of brightness unevenness defects in a more detailed manner, but also perform more accurate classification and scoring based on their time stability, spatial complexity and brightness change degree. Through the comprehensive analysis of these indices, the generated comprehensive score can clearly distinguish between minor, medium and severe defects, effectively improve the accuracy and robustness of detection, and significantly reduce the misjudgment and missed judgment caused by single-dimensional analysis. This method greatly improves the accuracy of liquid crystal panel quality evaluation and is particularly suitable for automated quality control of large-scale production lines.

[0016] (2) The present invention combines real-time monitoring with intelligent algorithms, and can dynamically adjust the detection time and parameter settings according to the actual situation of the area with uneven brightness problems, and intelligently deal with different types of uneven brightness defects. Unlike traditional methods that rely on fixed detection cycles and parameter settings, the present invention automatically optimizes the detection strategy by analyzing the spatial characteristics, brightness change amplitude and time stability of the uneven brightness area of ​​the liquid crystal panel in real time. When the system identifies a more serious uneven brightness defect, it automatically extends the detection time and improves the detection accuracy to ensure that such defects are fully identified; for minor defects, the system automatically shortens the detection time and adjusts the detection sensitivity to avoid redundant detection, significantly improving the detection efficiency. Through this intelligent dynamic adjustment strategy, the present invention can not only optimize the detection process, making the detection work more flexible, fast and accurate, but also effectively solve the problems of long-term detection and low-precision detection in traditional methods. Especially in the face of large-scale production and high-frequency detection, the solution of the present invention can significantly improve production efficiency and quality control accuracy, and provide a more efficient and intelligent quality detection solution in a rapidly changing manufacturing environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The present invention will be further described below in conjunction with the accompanying drawings.

[0018] Figure 1 It is a flowchart of the specific steps of a method for detecting uneven brightness defects of a liquid crystal panel according to the present invention; Figure 2 It is a flow chart of a detection system for uneven brightness defects of a liquid crystal panel in the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] See also Figure 1 As shown, the present invention is a method for detecting uneven brightness defects of a liquid crystal panel, comprising the following steps: S1: using a photometer to obtain an image of the liquid crystal panel in a display mode, and preprocessing the image to reduce noise and optimize contrast; S2: Identify and separate the area containing uneven brightness problem through image segmentation algorithm, analyze the brightness difference degree, spatial distribution and temporal stability of the area containing uneven brightness problem respectively, and calculate the spatial distribution characteristic index, brightness difference characteristic index and temporal stability characteristic index of the area containing uneven brightness problem according to the analysis results; S3: Comprehensively analyze the spatial distribution characteristic index, brightness difference characteristic index and temporal stability characteristic index of the extracted brightness unevenness problem area, and generate a comprehensive score for each brightness unevenness area based on the analysis results; S4: Each uneven brightness area is classified into slight defect, medium defect and severe defect according to the comprehensive score of each uneven brightness area; S5: Combined with real-time monitoring, the detection time and parameter settings of different types of defects are dynamically adjusted to achieve intelligent detection of different types of uneven brightness defects.

[0021] In S1, a photometer is used to obtain an image of the liquid crystal panel in a display mode, and the image is preprocessed to reduce noise and optimize contrast, specifically including: First, use a photometer to obtain image data of the LCD panel in a specific display mode. The photometer captures the brightness information of the LCD panel through a high-precision sensor to obtain the light intensity data of each pixel. In order to ensure the quality and accuracy of the obtained image, the photometer needs to set appropriate measurement parameters when performing image acquisition, including sampling frequency, exposure time, and image acquisition area. By adjusting these parameters, it is ensured that the acquired image reflects the actual brightness distribution of the LCD panel in the display mode.

[0022] Image preprocessing technology is used to process the collected images to reduce noise and optimize the image contrast to ensure more accurate subsequent analysis. The specific steps include: Noise removal: In the image of the LCD panel, noise may come from the interference of the equipment or environmental factors. Therefore, the image is firstly subjected to noise suppression using a filtering method, and the denoising technique uses Gaussian filtering. Gaussian filtering smoothes the image by weighted averaging the surrounding pixels, reducing the impact of random noise on the image.

[0023] Contrast optimization: LCD panel images have different brightness distributions, which makes it difficult to identify details in some areas. Therefore, histogram equalization technology is used to improve the contrast of the image, making the brightness values ​​more evenly distributed, thereby enhancing the visibility of uneven brightness areas in the image. Histogram equalization optimizes the contrast of the image by re-adjusting the brightness value range of the image, expanding the brightness difference, and ensuring that areas with uneven brightness can be clearly displayed.

[0024] In S2, the image segmentation algorithm is used to identify and separate the area containing the uneven brightness problem, and the brightness difference degree, spatial distribution and temporal stability of the area containing the uneven brightness problem are analyzed respectively. According to the analysis results, the spatial distribution characteristic index, brightness difference characteristic index and temporal stability characteristic index of the area containing the uneven brightness problem are calculated, which specifically include: The method of identifying and separating the area containing uneven brightness problem by using the image segmentation algorithm specifically includes: Through the adaptive threshold segmentation method, the brightness threshold is determined based on the brightness distribution of the image, and the area with brightness difference is extracted from the image; The image is processed based on a semantic segmentation network of deep learning, and a convolutional neural network is used to identify areas with uneven brightness. The convolutional neural network is trained by annotated data sets to learn the brightness characteristics of the liquid crystal panel for area recognition; The region growing algorithm is used to dynamically expand the region boundary according to the brightness difference of adjacent pixels and extract the problem area of ​​uneven brightness.

[0025] The above analysis respectively analyzes the brightness difference degree, spatial distribution and temporal stability of the brightness unevenness problem area, including: Analyze the degree of brightness difference in the area with uneven brightness problem, analyze the degree of brightness difference in the area with uneven brightness problem, and calculate the brightness difference characteristic index according to the amplitude of brightness fluctuation in the area with uneven brightness problem; Analyze the spatial distribution of the brightness unevenness problem area and extract the geometric characteristics of the spatial distribution, including the area, perimeter, length and width of the brightness unevenness problem area. According to the abnormality of the spatial distribution geometric characteristics, calculate the spatial distribution characteristic index; The temporal stability of the area with uneven brightness problem is analyzed, the stability of brightness change in the area with uneven brightness problem is analyzed, and the temporal stability characteristic index is calculated based on the analysis results, so as to evaluate the temporal stability of the area with uneven brightness problem.

[0026] The process of obtaining the brightness difference characteristic index is as follows: Apply discrete wavelet transform to the preprocessed image to decompose the image into low-frequency and high-frequency parts, including: Perform multi-level wavelet decomposition on the image to obtain the low-frequency component and high-frequency component of each layer; The calculation expression of each layer of wavelet decomposition is: ; ; in, is the low frequency component, is the high frequency component, and For low-pass and high-pass filters, represents the original image, Represents the pixel coordinates in the image; Calculate the brightness difference of each high-frequency component. The calculation expression is: ; in, Represents the pixel value of the high-frequency component, represents the mean value of high frequency components, represents the pixels in the high-frequency component, Represents the total number of pixels in the high-frequency component; Based on the calculated brightness difference, the brightness difference characteristic index of each high-frequency component is calculated, and the calculation expression is: ; in, is the total number of high-frequency subbands analyzed, represents the high frequency subband, Indicates The brightness difference of the high-frequency sub-bands, Indicates the brightness difference characteristic index of high-frequency components; The brightness difference characteristic index of the wavelet decomposition results of all levels is integrated to obtain the final brightness difference characteristic index. The calculation expression is: ; In the formula, represents the total number of wavelet decomposition layers, represents the wavelet decomposition layer, Represents the brightness difference characteristic index.

[0027] The process of obtaining the spatial distribution characteristic index is as follows: The area with uneven brightness is extracted, and each pixel is regarded as a node in the graph. Adjacent pixels are connected by edges, and the weight of the edge is represented by the difference in pixel brightness, including: Each edge in the graph connects two adjacent pixels, and the weight of the edge is Represents pixels and pixels The brightness difference between them is calculated as: ; in, Representation Node The brightness value of the corresponding pixel, Representation Node The brightness value of the corresponding pixel; Extract the geometric features of the area with uneven brightness, including: The area of ​​the region is calculated by the number of nodes in the uneven brightness area in the graph. The calculation expression is: ; in, is the number of nodes in the graph, Indicates the area of ​​uneven brightness; Calculate the perimeter of the area with uneven brightness. The calculation expression is: ; in, Indicates the perimeter of the area with uneven brightness, represents the edge set in the graph, Represents the indicator function. If the two nodes connected by the edge belong to the inside and outside of the region respectively, ,otherwise ; Calculate the aspect ratio of the area with uneven brightness. The calculation expression is: ; in, Represents the length of the circumscribed rectangle, Indicates the width of the bounding rectangle. Indicates the aspect ratio of the area with uneven brightness; According to the geometric characteristics of the uneven brightness area, the spatial distribution characteristic index is calculated comprehensively, and the calculation expression is: ; In the formula, Represents the spatial distribution characteristic index.

[0028] The process of obtaining the temporal stability characteristic index is as follows: Get the brightness values ​​of the area with uneven brightness at multiple time points to form a time series: ; in, Indicates the collection time point, Indicates the length of the time series; Through Fourier transformation, the time series Convert to frequency domain signal , the calculation expression is: ; in, represents the frequency domain signal, represents the imaginary unit, represents the logarithm of the base of a natural number, represents the frequency component; Calculate the amplitude spectrum of the frequency domain signal. The calculation expression is: ; in, Represents the frequency components in the frequency domain The corresponding amplitude; Find the main frequency and frequency energy of brightness changes by analyzing the amplitude spectrum; Among them, the main frequency is the frequency component corresponding to the maximum amplitude in the amplitude spectrum, which is recorded as ; Calculate the sum of the energies of all frequency components in the amplitude spectrum to obtain the total frequency energy. The calculation expression is: ; in, represents the total frequency energy, Indicates the maximum frequency of the frequency component; According to the calculation of total frequency energy and main frequency, the time stability characteristic index is calculated, and the calculation expression is: ; in, represents the temporal stability characteristic index, Representing time series The maximum amplitude, Indicates the main frequency.

[0029] In S3, the spatial distribution characteristic index, brightness difference characteristic index and temporal stability characteristic index of the extracted brightness unevenness problem area are comprehensively analyzed. According to the analysis results, a comprehensive score is generated for each brightness unevenness area, including: Obtain the spatial distribution characteristic index, brightness difference characteristic index and time stability characteristic index of the brightness unevenness problem area, perform normalized calculation processing on the spatial distribution characteristic index, brightness difference characteristic index and time stability characteristic index, and calculate the comprehensive score of the brightness unevenness area through the comprehensive score calculation expression; The comprehensive score calculation expression is: ; in, Represents the standardized value of the spatial distribution characteristic index, represents the normalized value of the brightness difference characteristic index, represents the normalized value of the temporal stability characteristic index, Indicates the overall rating.

[0030] According to the comprehensive score calculation expression, a comprehensive score is calculated for each area with uneven brightness problem, and a comprehensive score mark is made.

[0031] It should be noted that the comprehensive score reflects the degree of uneven defects in each area with uneven brightness problem, and the higher the value of the comprehensive score, the lower the degree of uneven defects in the corresponding area with uneven brightness problem.

[0032] In S4, each uneven brightness area is divided into slight defects, medium defects and severe defects according to the comprehensive score of each uneven brightness area, including: According to the comprehensive score calculation expression, the comprehensive score of each uneven brightness area is calculated respectively, and the comprehensive score of each uneven brightness area is compared with the first threshold value. If the comprehensive score is greater than or equal to the first threshold value, it means that the defect state of the corresponding area is slight, which is recorded as a slight defect; If the comprehensive score is less than the first threshold, the comprehensive score is compared with the second threshold. If the comprehensive score is less than or equal to the second threshold, it means that the defect state of the corresponding area is serious and is recorded as a serious defect; If the comprehensive score is greater than the second threshold and less than the first threshold, it means that the defect state of the corresponding area is medium, which is recorded as medium defect; The value of the first threshold is greater than the value of the second threshold.

[0033] It should be noted that the first standard threshold represents the lower limit of the regional defect state, the second standard threshold represents the upper limit of the regional defect state, and the regional defect state between the first standard threshold and the second standard threshold is a medium defect. The selection of the first standard threshold and the second standard threshold is based on the statistical analysis of historical data and is set according to the actual needs of the system and the error tolerance.

[0034] In S5, combined with real-time monitoring, the detection time and parameter settings of different types of defects are dynamically adjusted to achieve intelligent detection of different types of uneven brightness defects, including: By real-time monitoring of the changes in the uneven brightness area of ​​the LCD panel, the defect types are classified according to the brightness difference characteristics, spatial distribution characteristics and temporal stability characteristics.

[0035] For different types of uneven brightness defects, the detection time and detection parameters are adjusted in real time. By analyzing the complexity and stability of the defect area, the detection strategy is automatically optimized to improve detection efficiency and accuracy.

[0036] The time window and sensitivity parameters in the detection process are dynamically adjusted according to the characteristics of each uneven brightness defect. Specifically, when the system detects that the uneven brightness defect of the LCD panel in the area is a serious defect, the detection time is automatically extended and the detection accuracy is increased; if the uneven brightness defect of the LCD panel is a minor defect, the system automatically shortens the detection time and reduces redundant detection. When the system detects that the uneven brightness defect of the LCD panel in the area is a medium defect, the detection is performed according to the original detection time.

[0037] See also Figure 2 As shown, a system for detecting uneven brightness defects of a liquid crystal panel comprises: A data acquisition module, which uses a photometer to acquire an image of the liquid crystal panel in a display mode and preprocesses the image to reduce noise and optimize contrast; A region recognition module, which uses an image segmentation algorithm to identify and separate regions containing uneven brightness problems; A regional feature analysis module, which analyzes the brightness difference degree, spatial distribution and temporal stability of the brightness uneven problem area, and calculates the spatial distribution feature index, brightness difference feature index and temporal stability feature index of the brightness uneven problem area according to the analysis results; A comprehensive analysis module, which performs a comprehensive analysis on the spatial distribution characteristic index, the brightness difference characteristic index and the temporal stability characteristic index of the extracted brightness unevenness problem area, and generates a comprehensive score for each brightness unevenness area according to the analysis results; A defect recognition and classification module, wherein the defect recognition and classification module classifies each uneven brightness area into slight defects, medium defects and severe defects according to the comprehensive score of each uneven brightness area; A dynamic detection module combines real-time monitoring to dynamically adjust the detection time and parameter settings of different types of defects to achieve intelligent detection of different types of uneven brightness defects.

[0038] The working principle of the present invention is as follows: a photometer is used to obtain image data in the display mode of the liquid crystal panel, and the panel brightness information is captured by a high-precision sensor. The image is preprocessed using noise removal technology (such as Gaussian filtering) and contrast optimization (such as histogram equalization) to ensure the improvement of image quality and the clear display of areas with uneven brightness. Through image segmentation algorithms (such as adaptive threshold segmentation, deep learning semantic segmentation and regional growth algorithms), the areas with uneven brightness problems are identified and extracted, and their brightness differences, spatial distribution and temporal stability are analyzed respectively, and the spatial distribution characteristic index, brightness difference characteristic index and temporal stability characteristic index are calculated. Through the comprehensive analysis of these indices, a comprehensive score for each area with uneven brightness is generated, and the areas are divided into slight defects, medium defects and severe defects according to the scores.

[0039] The present invention also combines real-time monitoring means, can dynamically adjust the detection time and parameter settings, and perform intelligent detection for different types of uneven brightness defects. Specifically, the system will classify the defect type in real time according to the brightness difference characteristics, spatial distribution characteristics and time stability characteristics, automatically extend the detection time and improve the detection accuracy for serious defects; for minor defects, the detection time is automatically shortened to avoid redundant detection and improve detection efficiency. Medium defects are processed according to the original detection time, thereby optimizing the detection process and enhancing the flexibility and intelligence level of detection. Through this method, efficient detection and accurate classification of uneven brightness problems of liquid crystal panels can be achieved, and the intelligence level of product quality monitoring can be improved.

[0040] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0041] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0042] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0043] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for detecting uneven brightness defects of a liquid crystal panel, characterized in that: The following steps are involved: S1: using a photometer to obtain an image of the liquid crystal panel in a display mode, and preprocessing the image to reduce noise and optimize contrast; S2: Identify and separate the area containing uneven brightness problem through image segmentation algorithm, analyze the brightness difference degree, spatial distribution and temporal stability of the area containing uneven brightness problem respectively, and calculate the spatial distribution characteristic index, brightness difference characteristic index and temporal stability characteristic index of the area containing uneven brightness problem according to the analysis results; S3: Comprehensively analyze the spatial distribution characteristic index, brightness difference characteristic index and temporal stability characteristic index of the extracted brightness unevenness problem area, and generate a comprehensive score for each brightness unevenness area based on the analysis results; S4: Each uneven brightness area is classified into slight defect, medium defect and severe defect according to the comprehensive score of each uneven brightness area; S5: Combined with real-time monitoring, the detection time and parameter settings of different types of defects are dynamically adjusted to achieve intelligent detection of different types of uneven brightness defects.

2. A method for detecting uneven brightness defects of a liquid crystal panel according to claim 1, characterized in that: The method of identifying and separating the area containing uneven brightness problem by using the image segmentation algorithm specifically includes: Through the adaptive threshold segmentation method, the brightness threshold is determined based on the brightness distribution of the image, and the area with brightness difference is extracted from the image; The image is processed based on a semantic segmentation network of deep learning, and a convolutional neural network is used to identify areas with uneven brightness. The convolutional neural network is trained by annotated data sets to learn the brightness characteristics of the liquid crystal panel for area recognition; The region growing algorithm is used to dynamically expand the region boundary according to the brightness difference of adjacent pixels and extract the problem area of ​​uneven brightness.

3. The method for detecting uneven brightness defects of a liquid crystal panel according to claim 1, characterized in that: The analysis of the brightness difference degree, spatial distribution and temporal stability of the brightness unevenness problem area specifically includes: Analyze the degree of brightness difference in the area with uneven brightness problem, analyze the degree of brightness difference in the area with uneven brightness problem, and calculate the brightness difference characteristic index according to the amplitude of brightness fluctuation in the area with uneven brightness problem; Analyze the spatial distribution of the brightness unevenness problem area and extract the geometric characteristics of the spatial distribution, including the area, perimeter, length and width of the brightness unevenness problem area. According to the abnormality of the spatial distribution geometric characteristics, calculate the spatial distribution characteristic index; The temporal stability of the area with uneven brightness problem is analyzed, the stability of brightness change in the area with uneven brightness problem is analyzed, and the temporal stability characteristic index is calculated based on the analysis results, so as to evaluate the temporal stability of the area with uneven brightness problem.

4. The method for detecting uneven brightness defects of a liquid crystal panel according to claim 3, characterized in that: The process of obtaining the brightness difference characteristic index is as follows: Apply discrete wavelet transform to the preprocessed image to decompose the image into low-frequency and high-frequency parts, including: Perform multi-level wavelet decomposition on the image to obtain the low-frequency component and high-frequency component of each layer; The calculation expression of each layer of wavelet decomposition is: ; ; in, is the low frequency component, is the high frequency component, and For low-pass and high-pass filters, represents the original image, Represents the pixel coordinates in the image; Calculate the brightness difference of each high-frequency component. The calculation expression is: ; in, Represents the pixel value of the high-frequency component, represents the mean value of high frequency components, represents the pixels in the high-frequency component, Represents the total number of pixels in the high-frequency component; Based on the calculated brightness difference, the brightness difference characteristic index of each high-frequency component is calculated, and the calculation expression is: ; in, is the total number of high-frequency subbands analyzed, represents the high frequency subband, Indicates The brightness difference of the high-frequency sub-bands, Indicates the brightness difference characteristic index of high-frequency components; The brightness difference characteristic index of the wavelet decomposition results of all levels is integrated to obtain the final brightness difference characteristic index. The calculation expression is: ; In the formula, represents the total number of wavelet decomposition layers, represents the wavelet decomposition layer, Represents the brightness difference characteristic index.

5. The method for detecting uneven brightness defects of a liquid crystal panel according to claim 3, characterized in that: The process of obtaining the spatial distribution characteristic index is as follows: The uneven brightness area is extracted, and each pixel is regarded as a node in the graph. Adjacent pixels are connected by edges, and the weight of the edge is represented by the difference in pixel brightness, including: Each edge in the graph connects two adjacent pixels, and the weight of the edge is Represents pixels and pixels The brightness difference between them is calculated as: ; in, Representation Node The brightness value of the corresponding pixel, Representation Node The brightness value of the corresponding pixel; Extract the geometric features of the area with uneven brightness, including: The area of ​​the region is calculated by the number of nodes in the uneven brightness area in the graph. The calculation expression is: ; in, is the number of nodes in the graph, Indicates the area of ​​uneven brightness; Calculate the perimeter of the area with uneven brightness. The calculation expression is: ; in, Indicates the perimeter of the area with uneven brightness, represents the edge set in the graph, Represents the indicator function. If the two nodes connected by the edge belong to the inside and outside of the region respectively, ,otherwise ; Calculate the aspect ratio of the area with uneven brightness. The calculation expression is: ; in, Represents the length of the circumscribed rectangle, Indicates the width of the bounding rectangle. Indicates the aspect ratio of the area with uneven brightness; According to the geometric characteristics of the uneven brightness area, the spatial distribution characteristic index is calculated comprehensively, and the calculation expression is: ; In the formula, Represents the spatial distribution characteristic index.

6. The method for detecting uneven brightness defects of a liquid crystal panel according to claim 3, characterized in that: The process of obtaining the temporal stability characteristic index is as follows: Get the brightness values ​​of the area with uneven brightness at multiple time points to form a time series: ; in, Indicates the collection time point, Indicates the length of the time series; Through Fourier transformation, the time series Convert to frequency domain signal , the calculation expression is: ; in, represents the frequency domain signal, represents the imaginary unit, represents the logarithm of the base of a natural number, represents the frequency component; Calculate the amplitude spectrum of the frequency domain signal. The calculation expression is: ; in, Represents the frequency components in the frequency domain The corresponding amplitude; Find the main frequency and frequency energy of brightness changes by analyzing the amplitude spectrum; Among them, the main frequency is the frequency component corresponding to the maximum amplitude in the amplitude spectrum, which is recorded as ; Calculate the sum of the energies of all frequency components in the amplitude spectrum to obtain the total frequency energy. The calculation expression is: ； in, represents the total frequency energy, Indicates the maximum frequency of the frequency component; According to the calculation of total frequency energy and main frequency, the time stability characteristic index is calculated, and the calculation expression is: ； in, represents the temporal stability characteristic index, Representing time series The maximum amplitude, Indicates the main frequency.

7. The method for detecting uneven brightness defects of a liquid crystal panel according to claim 1, characterized in that: The spatial distribution characteristic index, brightness difference characteristic index and time stability characteristic index of the extracted brightness uneven problem area are comprehensively analyzed, and a comprehensive score is generated for each brightness uneven area according to the analysis results, specifically including: Obtain the spatial distribution characteristic index, brightness difference characteristic index and time stability characteristic index of the brightness unevenness problem area, perform normalized calculation processing on the spatial distribution characteristic index, brightness difference characteristic index and time stability characteristic index, and calculate the comprehensive score of the brightness unevenness area through the comprehensive score calculation expression; The comprehensive score calculation expression is: ； in, Represents the standardized value of the spatial distribution characteristic index, represents the normalized value of the brightness difference characteristic index, represents the normalized value of the temporal stability characteristic index, Indicates the overall rating.

8. The method for detecting uneven brightness defects of a liquid crystal panel according to claim 1, characterized in that: According to the comprehensive score of each uneven brightness area, each uneven brightness area is divided into slight defects, medium defects and severe defects, specifically including: According to the comprehensive score calculation expression, the comprehensive score of each uneven brightness area is calculated respectively, and the comprehensive score of each uneven brightness area is compared with the first threshold value. If the comprehensive score is greater than or equal to the first threshold value, it means that the defect state of the corresponding area is slight, which is recorded as a slight defect; If the comprehensive score is less than the first threshold, the comprehensive score is compared with the second threshold. If the comprehensive score is less than or equal to the second threshold, it means that the defect state of the corresponding area is serious and is recorded as a serious defect; If the comprehensive score is greater than the second threshold and less than the first threshold, it means that the defect state of the corresponding area is medium, which is recorded as medium defect; The value of the first threshold is greater than the value of the second threshold.

9. The method for detecting uneven brightness defects of a liquid crystal panel according to claim 1, characterized in that: The above-mentioned real-time monitoring is combined to dynamically adjust the detection time and parameter settings of different types of defects to achieve intelligent detection of different types of uneven brightness defects, specifically including: By monitoring the changes in the uneven brightness area of ​​the LCD panel in real time, the defect types are classified according to the brightness difference characteristics, spatial distribution characteristics and time stability characteristics; When the system detects that the uneven brightness defect of the LCD panel in the area is a serious defect, it will automatically extend the detection time and increase the detection accuracy; if the uneven brightness defect of the LCD panel is a minor defect, the system will automatically shorten the detection time and reduce redundant detection. When the system detects that the uneven brightness defect of the LCD panel in the area is a medium defect, it will be detected according to the original detection time.

10. A system for detecting uneven brightness defects of a liquid crystal panel, characterized in that: A method for detecting uneven brightness defects of a liquid crystal panel as claimed in any one of claims 1 to 9, comprising: A data acquisition module, which uses a photometer to acquire an image of the liquid crystal panel in a display mode and preprocesses the image to reduce noise and optimize contrast; A region recognition module, which uses an image segmentation algorithm to identify and separate regions containing uneven brightness problems; A regional feature analysis module, which analyzes the brightness difference degree, spatial distribution and temporal stability of the brightness uneven problem area, and calculates the spatial distribution feature index, brightness difference feature index and temporal stability feature index of the brightness uneven problem area according to the analysis results; A comprehensive analysis module, which performs a comprehensive analysis on the spatial distribution characteristic index, the brightness difference characteristic index and the temporal stability characteristic index of the extracted brightness unevenness problem area, and generates a comprehensive score for each brightness unevenness area according to the analysis results; A defect recognition and classification module, wherein the defect recognition and classification module classifies each uneven brightness area into slight defects, medium defects and severe defects according to the comprehensive score of each uneven brightness area; A dynamic detection module combines real-time monitoring to dynamically adjust the detection time and parameter settings of different types of defects to achieve intelligent detection of different types of uneven brightness defects.

Images