A method and system for inspecting the edge grinding quality of mobile phone glass covers.

Abstract

This invention relates to the field of image data processing technology, specifically to a method and system for detecting the edge polishing quality of mobile phone glass covers. The method includes: capturing an image of the mobile phone glass cover to be inspected; filtering the image using an improved anisotropic diffusion filtering algorithm; and performing edge detection on the filtered image to obtain multiple edge contours. If the number of edge contours exceeds a set threshold, the image is deemed to be of substandard quality. This invention solves the problem of low accuracy in edge polishing quality detection of mobile phone glass covers using existing anisotropic diffusion filtering algorithms.

Description

Technical Field

[0001] This invention relates to the field of image data processing technology. More specifically, this invention relates to a method and system for detecting the edge polishing quality of mobile phone glass covers. Background Technology

[0002] In modern consumer electronics, the mobile phone glass cover is a key component of the display module, and its appearance quality and structural integrity directly affect the overall aesthetics, touch sensitivity, and user experience. Especially in high-end smartphones, with increasingly stringent requirements for appearance craftsmanship, the edges of the glass cover often undergo complex and precise polishing processes to achieve smooth edge transitions, prevent cuts, and enhance the visual effect of the integrated design. However, in actual production, due to factors such as fluctuations in polishing process precision, insufficient equipment stability, and uneven microstructure of raw materials, defects such as micro-cracks, chipping, over-polishing, and insufficient polishing are easily generated at the edges of the glass cover.

[0003] Traditional anisotropic diffusion filtering algorithms are used in tasks such as edge recognition, texture analysis, and defect detection due to their combination of image denoising and edge protection. This algorithm constructs a pixel diffusion model within a local region, enabling selective spatial smoothing of the image. This allows for the suppression of random noise interference while preserving key structural information, making it particularly suitable for image preprocessing before high-contrast edge detection.

[0004] However, in practical engineering applications, traditional anisotropic diffusion filtering algorithms generally use fixed diffusion control parameters, i.e., the diffusion function... The value remains unchanged. The value determines the intensity of the diffusion function's response to differences in grayscale between pixels. Setting the value too high will cause the image edges to be overly smoothed, weakening or even blurring edge information; while Setting the value too small restricts the diffusion process, making it impossible to effectively suppress background noise. This can lead to false contours or excessive detail retention in the texture area, resulting in low accuracy of subsequent edge detection. Summary of the Invention

[0005] To address the problem of low accuracy mentioned in the background art, the present invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a method for detecting the edge polishing quality of a mobile phone glass cover, comprising: capturing an image of a mobile phone glass cover to be inspected; filtering the image of the mobile phone glass cover to be inspected using an improved anisotropic diffusion filtering algorithm, and performing edge detection on the filtered image of the mobile phone glass cover to be inspected to obtain multiple edge contours; if the number of edge contours is greater than a set threshold, the image of the mobile phone glass cover to be inspected is determined to be of unqualified quality; wherein, the improved anisotropic diffusion filtering algorithm includes Value, the Value and initial The value is positively correlated with the degree of anomaly of each pixel, and negatively correlated with the edge sensitivity of each pixel.

[0007] The aforementioned technical solution introduces a dynamic adjustment mechanism based on pixel anomaly severity and edge sensitivity, enabling intelligent optimization of anisotropic diffusion filter parameters. This allows the filtering process to adaptively adjust the smoothing intensity for different regions, effectively enhancing noise suppression capabilities in abnormal areas while protecting the integrity of critical edge structures. Combined with improved image edge detection, a set threshold is used as the quality judgment criterion by statistically analyzing the number of edge contours from multiple qualified samples, thus improving the accuracy of defect identification.

[0008] Furthermore, pixels of value for: , For the initial value, For normalization function, For pixels The degree of abnormality, For the natural constant An exponential function with base 0. For pixels Edge sensitivity.

[0009] The above technical solution combines the degree of pixel anomaly with edge sensitivity to dynamically adjust key parameters in diffusion filtering, thereby achieving adaptive allocation of filtering intensity. This enhances diffusion strength in abnormal areas to effectively suppress noise, while significantly weakening diffusion through an exponential decay mechanism in edge-sensitive areas, thus effectively protecting the details, edges, and structural features in the image.

[0010] Further, the degree of abnormality of any target pixel in the mobile phone glass cover image to be detected is determined as follows: The grayscale value of the target pixel and the grayscale values ​​of all pixels within a defined neighborhood centered on the target pixel are obtained; for each pixel within the defined neighborhood, the absolute value of the difference between its grayscale value and that of the target pixel, and the absolute value of the difference between its grayscale value and the average grayscale value of all pixels within the defined neighborhood, are calculated, and these two values ​​are multiplied together; the multiplication results corresponding to all pixels within the defined neighborhood are accumulated to obtain the degree of abnormality of the target pixel, where the target pixel is any pixel in the mobile phone glass cover image to be detected.

[0011] The above technical solution constructs a weighted evaluation method for local gray-level anomalies by multiplying the gray-level difference between the target pixel and its neighboring pixels by the deviation of the neighboring pixels from the local mean. This method effectively enhances the response capability to problems such as excessive edge blurring, detail disorder, and local structural anomalies. When there are obvious gray-level jumps or abrupt changes in the neighborhood, the index value will increase significantly, reflecting poor local stability or potential defects in the area where the current pixel is located.

[0012] Further, the edge sensitivity is as follows: The gradient magnitude of the target pixel and the gradient magnitudes of all pixels within a defined neighborhood centered on the target pixel are obtained; the grayscale variance of all pixels within the defined neighborhood and the grayscale variance of all pixels in the mobile phone glass cover image to be detected are calculated; for each pixel within the defined neighborhood, the absolute value of the difference between its gradient magnitude and that of the target pixel is calculated, and the absolute values ​​of the differences between the gradient magnitudes of all pixels within the defined neighborhood and that of the target pixel are accumulated; the ratio of the grayscale variance of all pixels within the defined neighborhood to the grayscale variance of all pixels in the mobile phone glass cover image to be detected is calculated, and the ratio is multiplied by the accumulated result to obtain the edge sensitivity of the target pixel, where the target pixel is any pixel in the mobile phone glass cover image to be detected.

[0013] The aforementioned technical solution integrates local gray-level variance and gradient magnitude difference to construct a sensitivity metric that comprehensively reflects the complexity of pixel edge features, effectively improving the identification capability of key structural regions in images. Specifically, local gray-level variance measures the intensity of gray-level fluctuations in the region surrounding the current pixel, reflecting the richness of texture or potential abnormal disturbances in that region; while the accumulation of gradient magnitude differences further characterizes the drastic changes in local structural edges, thereby enhancing the response capability to edge polishing defects such as detailed edges, cracks, and depressions.

[0014] Furthermore, an image of the glass cover of the mobile phone to be tested is captured using a CCD camera or a CMOS camera.

[0015] Furthermore, the Canny edge detection algorithm is used to perform edge detection on the image of the mobile phone glass cover to be detected.

[0016] Furthermore, the image of the glass cover of the mobile phone to be tested is processed into grayscale.

[0017] Furthermore, the set threshold is the average number of edge contours among multiple qualified mobile phone glass cover images.

[0018] The above technical solution uses the number of edge contours in multiple qualified glass cover images for statistical modeling, which can fully reflect the distribution of edge structure features of the image under stable process conditions and form a representative reference benchmark. Thus, in actual testing, when the number of edges in the image under test is abnormally increased due to cracks, chipping, foreign objects or other defects, it can be compared with the mean value to quickly determine whether the image quality is abnormal, thereby improving the scientificity and interpretability of the test results.

[0019] Furthermore, it also includes triggering an alarm to prompt manual re-inspection if the image quality of the glass cover of the mobile phone to be inspected is determined to be unqualified.

[0020] In a second aspect, the present invention provides a system for detecting the edge polishing quality of a mobile phone glass cover, comprising a memory and a processor. The memory stores computer program instructions, which, when executed by the processor, implement the edge polishing quality detection method for a mobile phone glass cover as described above.

[0021] The beneficial effects of this invention are as follows:

[0022] This invention utilizes an improved anisotropic diffusion filtering algorithm, dynamically adjusting diffusion parameters based on the local anomaly level and edge sensitivity of pixels. This effectively suppresses image noise while preserving key edge information, enhancing the ability to identify subtle grinding defects. Furthermore, in the edge detection stage, an adaptive threshold based on historical qualified sample statistics is introduced as a judgment criterion, which not only enhances the robustness of the system but also improves its adaptability to process fluctuations, significantly improving the accuracy of detection. Attached Figure Description

[0023] Figure 1 This is a schematic flowchart illustrating a method for detecting the edge polishing quality of a mobile phone glass cover according to an embodiment of the present invention.

[0024] Figure 2 This is a schematic block diagram illustrating the structure of a mobile phone glass cover edge polishing quality inspection system according to an embodiment of the present invention. Detailed Implementation

[0025] An embodiment of a method for detecting the edge polishing quality of a mobile phone glass cover.

[0026] like Figure 1 The flowchart shown is a method for inspecting the edge polishing quality of a mobile phone glass cover according to an embodiment of the present invention, which includes the following steps:

[0027] S1: Capture an image of the glass cover of the mobile phone to be tested.

[0028] In a preferred embodiment, high-resolution images of the mobile phone glass cover to be inspected are first acquired using an industrial-grade CCD camera or a high-performance CMOS camera. The core of this step lies in leveraging the advantages of CCD or CMOS image sensors in image clarity, response speed, and noise control to accurately capture the microstructural features of the glass cover's edges and surface. Especially in the glass cover edge area, where fine polishing processes are involved, micro-defects such as chipping, breakage, uneven polishing, and cracks are easily generated. If the accuracy of the image acquisition stage is insufficient, subsequent inspection processes will struggle to accurately identify these critical defects, thus affecting the effectiveness of the inspection. Therefore, using a high-precision camera for image acquisition not only ensures image quality but also improves the lower limit sensitivity and recognition resolution of defect detection, enhancing the ability to perceive low-contrast and low-brightness defects.

[0029] Next, to further simplify the image data structure, highlight the grayscale features of the target region, and suppress irrelevant color interference, the acquired original image is converted to grayscale. The main purpose of grayscale conversion is to map the original image from a multi-channel color space (such as RGB) to a single-channel grayscale space, eliminating redundant calculations and color interference caused by the color dimension, making the image content clearer and more prominent in terms of structure, contours, and brightness variations. This conversion helps subsequent image processing algorithms, especially edge detection, texture analysis, and contour extraction, to operate more efficiently.

[0030] Furthermore, due to the inherent transparency and reflectivity of glass, the acquired raw images are often susceptible to various factors such as ambient lighting, reflected highlights, and background complexity, leading to blurred edge contours or localized texture distortion. By converting the image to grayscale, these unstable factors can be smoothed out to some extent, making the image's brightness distribution more consistent with the brightness gradient and texture stability requirements of subsequent feature extraction algorithms. This improves the system's ability to distinguish between edge transition areas and defect regions, enhancing the accuracy and robustness of the detection results.

[0031] S2: The improved anisotropic diffusion filtering algorithm is used to filter the image of the mobile phone glass cover to be detected.

[0032] In a preferred embodiment, the improved anisotropic diffusion filtering algorithm includes: Value, pixel of value for: , For the initial value, For normalization function, For pixels The degree of abnormality, For the natural constant An exponential function with base 0. For pixels Edge sensitivity.

[0033] By introducing an adaptive adjustment mechanism that combines anomaly level and edge sensitivity, dynamic values ​​are assigned to key parameters in anisotropic diffusion filtering, thereby better protecting important edge structures while smoothing the image. Specifically, when a pixel has a high anomaly level and low edge sensitivity, its diffusion control parameters are increased accordingly, which helps to enhance the suppression of local noise; while when edge sensitivity is high, the diffusion intensity is suppressed by the rapid decay characteristic of the exponential function, thus effectively preventing over-smoothing of edge regions. Overall, the above scheme improves denoising performance while maintaining image structural clarity, achieving a dynamic balance between image smoothness and edge preservation capability, and significantly improving the accuracy and robustness of subsequent image analysis and feature extraction tasks.

[0034] abnormality for: , For pixels grayscale value, In pixels Pixels within the set neighborhood of the center grayscale value, In pixels The average grayscale value of all pixels within a neighborhood of the center is set. The total number of pixels within the defined neighborhood is set. By assigning the grayscale difference between the target pixel and its neighboring pixels as a weight to the degree to which the neighboring pixels deviate from the regional mean, a differentiated weighted processing of the degree of local anomalies is achieved. This allows neighboring pixels with larger grayscale differences from the target pixel to occupy a higher weight in the overall evaluation, thus more sensitively reflecting the non-uniformity and abnormal fluctuations in local image regions. This weighting mechanism can effectively highlight areas with drastic grayscale changes, structural discontinuities, or abrupt texture changes, enhancing the ability to identify image anomalies such as edge anomalies, flaw features, and noise interference. At the same time, it avoids the over-amplification of weak changes in uniform regions, thereby improving the accuracy and robustness of anomaly detection.

[0035] Edge sensitivity for: , In pixels The grayscale variance of all pixels within a neighborhood centered on the given value. Let V be the grayscale variance of the entire image of the glass cover of the mobile phone to be tested. For pixels gradient magnitude, In pixels Pixels within the set neighborhood of the center gradient magnitude, The total number of pixels within the defined neighborhood. These are the preset hyperparameters.

[0036] By jointly considering the degree of local gray-level fluctuation and the magnitude of gradient change, a refined model of edge region sensitivity is achieved. The ratio of local gray-level variance to the overall gray-level variance of the entire image is used to measure the contrast intensity of the pixel against the global background, while the accumulation of gradient magnitude differences within the neighborhood reflects the severity of local structural changes. This fusion of dual features effectively enhances the recognition ability of edge transition regions, contour change regions, and regions with abrupt changes in detail texture. It is particularly suitable for detecting edge anomalies in glass covers caused by uneven polishing, microcracks, or minor defects. At the same time, the normalization mechanism introduced by the global variance also improves the adaptability and stability under different image contrast conditions, thereby ensuring sensitive recognition while avoiding misjudgment.

[0037] S3: Perform edge detection on the filtered image of the mobile phone glass cover to be tested to obtain multiple edge contours. If the number of edge contours is greater than the set threshold, the image of the mobile phone glass cover to be tested is deemed to be of unqualified quality.

[0038] In a preferred embodiment, after anisotropic diffusion filtering of the image of the mobile phone glass cover to be tested, edge detection is further performed to extract significant edge contour information from the image, which is used to evaluate the polishing quality of the cover edge region and its overall structural integrity. The classic Canny edge detection algorithm is preferably used in this process, as it has high accuracy and stability in gradient calculation, double threshold connection, and non-maximum suppression. It can effectively identify grayscale abrupt change regions in the image caused by polishing defects, cracks, or edge damage, thereby obtaining complete and representative edge contour data.

[0039] After edge contour extraction, the number of all edge contours in the image is counted and compared with a preset judgment threshold to determine whether the glass cover corresponding to the image has quality abnormalities. This threshold is calculated by averaging the number of edge contours in multiple historically acquired quality-compliant sample images, using this average as a benchmark value to quantify the edge structural complexity of the current detection image. The design logic of this scheme is that when defects such as microcracks, uneven polishing, or broken edges exist on the glass cover surface, the edge detection results usually show an abnormally increased number of edge contour features. Therefore, by setting a reasonable threshold for the number of contours, such structural abnormalities can be effectively identified, and rapid screening of overall quality can be achieved with low computational cost.

[0040] Furthermore, when the test results determine that the current image quality is unqualified, an alarm mechanism will be automatically triggered to issue a clear prompt to the operator, thereby initiating a manual re-inspection process. This prevents potentially defective products from flowing into subsequent production stages or being shipped as finished products. In this way, while ensuring the level of automation in the testing, manual verification is introduced as a supplementary means to build a more robust, accurate, and traceable quality inspection process.

[0041] This invention utilizes an improved anisotropic diffusion filtering algorithm to enhance the identification of local anomaly regions in images while effectively suppressing noise and preserving edge structure, ensuring the accuracy and continuity of edge feature extraction. By constructing a diffusion adjustment factor related to the degree of image grayscale anomaly and edge sensitivity, pixel-level adaptive control of the filtering intensity is achieved, enabling the system to sensitively detect subtle defects and reduce false responses to complex backgrounds. The edge contour quantity threshold obtained through statistical modeling is used for quality judgment, significantly improving the objectivity and generalization ability of the detection, effectively enhancing the automation level and accuracy of edge polishing quality inspection, and providing strong technical support for the quality control of industrial-grade mobile phone glass covers.

[0042] An embodiment of a mobile phone glass cover edge polishing quality inspection system:

[0043] like Figure 2 As shown, a structural block diagram of an edge polishing quality inspection system for a mobile phone glass cover according to an embodiment of the present invention includes a processor and a memory.

[0044] This invention also provides a system for inspecting the edge polishing quality of mobile phone glass covers. For example... Figure 2 As shown, the system includes a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a method for detecting the edge polishing quality of a mobile phone glass cover according to the present invention.

[0045] The edge polishing quality inspection system for mobile phone glass covers also includes other components well known to those skilled in the art, such as communication interfaces. Their settings and functions are known in the art and will not be described in detail here.

[0046] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented by computer-readable / executable instructions stored or otherwise maintained on such a computer-readable medium.

[0047] In the description of this specification, "multiple" or "several" means at least two, such as two, three or more, unless otherwise explicitly specified.

[0048] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.

Claims

1. A method for inspecting the edge grinding quality of a mobile phone glass cover, characterized in that, include: Capture an image of the glass cover of the mobile phone to be tested; An improved anisotropic diffusion filtering algorithm is used to filter the image of the mobile phone glass cover to be tested, and edge detection is performed on the filtered image to obtain multiple edge contours. If the number of edge contours is greater than a set threshold, the image of the mobile phone glass cover to be tested is determined to be of unqualified quality. Among them, the improved anisotropic diffusion filtering algorithm includes Value, the Value and initial The value is positively correlated with the degree of anomaly of each pixel, and negatively correlated with the edge sensitivity of each pixel. When a pixel has a high degree of anomaly and low edge sensitivity, its diffusion control parameter will be increased accordingly, which is beneficial to enhance the suppression of local noise; while when the edge sensitivity is high, the diffusion intensity is suppressed by the rapid decay characteristic of the exponential function, thereby effectively preventing excessive smoothing of the edge region. pixel of value for: , For the initial value, For normalization function, For pixels The degree of abnormality, For the natural constant An exponential function with base 0. For pixels Edge sensitivity; The edge sensitivity is defined as follows: The gradient magnitude of the target pixel and the gradient magnitudes of all pixels within a defined neighborhood centered on the target pixel are obtained; the grayscale variance of all pixels within the defined neighborhood and the grayscale variance of all pixels in the mobile phone glass cover image to be tested are calculated; for each pixel within the defined neighborhood, the absolute value of the difference between its gradient magnitude and that of the target pixel is calculated, and the absolute values ​​of the differences between the gradient magnitudes of all pixels within the defined neighborhood and that of the target pixel are accumulated; the ratio of the grayscale variance of all pixels within the defined neighborhood to the grayscale variance of all pixels in the mobile phone glass cover image to be tested is calculated, and the ratio is multiplied by the accumulated result to obtain the edge sensitivity of the target pixel, where the target pixel is any pixel in the mobile phone glass cover image to be tested.

2. The method for detecting the edge grinding quality of a mobile phone glass cover according to claim 1, characterized in that, The degree of abnormality is defined as follows: obtaining the grayscale value of the target pixel and the grayscale values ​​of all pixels within a defined neighborhood centered on the target pixel; For each pixel within the defined neighborhood, calculate the absolute value of the difference between its gray value and the target pixel's gray value, and the absolute value of the difference between its gray value and the average gray value of all pixels within the defined neighborhood, and multiply the two together. The multiplication results of all pixels in the defined neighborhood range are summed to obtain the degree of abnormality of the target pixel, which is any pixel in the image of the mobile phone glass cover to be detected.

3. The method for detecting the edge grinding quality of a mobile phone glass cover according to claim 1, characterized in that, Use a CCD or CMOS camera to capture images of the glass cover of the mobile phone to be tested.

4. The method for detecting the edge grinding quality of a mobile phone glass cover according to claim 1, characterized in that, The Canny edge detection algorithm is used to perform edge detection on the image of the glass cover of the mobile phone to be tested.

5. The method for detecting the edge grinding quality of a mobile phone glass cover according to claim 1, characterized in that, The image of the glass cover of the mobile phone to be tested is converted to grayscale.

6. The method for detecting the edge grinding quality of a mobile phone glass cover according to claim 1, characterized in that, The set threshold is the average number of edge contours in multiple qualified mobile phone glass cover images.

7. The method for detecting the edge grinding quality of a mobile phone glass cover according to claim 1, characterized in that, It also includes triggering an alarm to prompt manual re-inspection if the image quality of the glass cover of the mobile phone to be inspected is deemed unqualified.

8. A system for inspecting the edge grinding quality of a mobile phone glass cover, characterized in that, The device includes a memory and a processor. The memory stores computer program instructions, which, when executed by the processor, implement the edge polishing quality detection method for mobile phone glass covers as described in any one of claims 1 to 7.

Images