Bathroom towel rack plug detection method and system based on AI visual information

Abstract

The invention discloses a bathroom towel rack plug detection method and system based on AI visual information, and relates to the technical field of visual identification, and the method comprises the steps: obtaining bathroom towel rack plug images at different angles, and extracting a bathroom towel rack plug feature image; based on a region analysis algorithm, analyzing the bathroom towel rack plug feature image, and identifying a defect region of the bathroom towel rack plug; identifying the defect area by using a defect division algorithm to obtain the defect type of the bathroom towel rack plug; and in combination with the defect type, automatically scoring the quality of the bathroom towel rack plug by using a scoring algorithm, and generating a defect report. According to the method, area analysis and a defect division algorithm are combined, defect areas and types are automatically identified, careless omission and misjudgment in detection are avoided, the defect types are further distinguished by using the defect division algorithm, the product quality can be accurately evaluated, and the defective product rate is reduced.

Description

Technical Field

[0001] The present invention relates to the field of visual recognition technology, and in particular to a method and system for detecting blockages in bathroom towel racks based on AI visual information. Background Art

[0002] With the continuous advancement of artificial intelligence technology, the application of AI visual information processing technology in industrial quality inspection has gradually become popular. By integrating image acquisition, deep learning models and feature recognition algorithms, AI vision can simulate the human visual system to perform multi-dimensional analysis of images. It has significant advantages such as non-contact, high precision, strong real-time performance, and adaptability to complex environments. It is particularly suitable for product appearance inspection with complex defect morphology and multi-scale changes. In the inspection scenario of bathroom towel rack plugs, since the surface of the plug is often accompanied by reflections, curved structures and various minor defects such as cracks, scratches, indentations and pores, AI vision technology can accurately identify and classify various defect areas through image recognition and deep feature extraction, improve detection efficiency and stability, and provide key support for the quality control of bathroom plugs and subsequent intelligent process optimization.

[0003] However, the existing technology has limited accuracy in segmenting and identifying defective areas, making it difficult to effectively distinguish similar or overlapping defect types, resulting in inaccurate classification results. It is also difficult to fully capture the surface details of the plug, increasing omissions and misjudgments in detection. At the same time, there is a lack of a quantitative scoring mechanism for the severity of defects, making it difficult to use the detection results for subsequent production optimization. The defect data cannot form an effective feedback loop, which limits the quality inspection link's ability to intelligently control and improve the production process.

[0004] Currently, no effective solutions have been proposed for the problems in related technologies. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the present invention proposes a bathroom towel rack blockage detection method and system based on AI visual information, which solves the problems raised in the above background technology, such as the limited accuracy of the existing segmentation and identification of defect areas, the inconvenience in effectively distinguishing similar or overlapping defect types, resulting in inaccurate classification results, and the inconvenience in fully capturing the surface details of the blockage, thereby increasing omissions and misjudgments in detection. At the same time, there is a lack of a quantitative scoring mechanism for the severity of defects, and the detection results are difficult to use for subsequent production optimization. The defect data cannot form an effective feedback loop, which limits the quality inspection link's ability to intelligently control and improve the production process.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0007] According to one aspect of the present invention, a method for detecting blockages in bathroom towel racks based on AI visual information is provided, comprising:

[0008] S1. Acquire images of bathroom towel rack plugs at different angles and extract feature images of bathroom towel rack plugs;

[0009] S2. Analyze the characteristic image of the bathroom towel rack plug based on the regional analysis algorithm to identify the defective area of the bathroom towel rack plug;

[0010] S3. Using a defect segmentation algorithm to identify the defect area, and obtaining the defect type of the bathroom towel rack plug;

[0011] S4. Based on the defect type, the scoring algorithm is used to automatically score the quality of the bathroom towel rack plugs and generate a defect report. Based on the defect report, the production process is adjusted in real time and the production process is optimized to reduce the defective product rate.

[0012] Furthermore, images of the bathroom towel rack plug are obtained at different angles, and feature images of the bathroom towel rack plug are extracted, including:

[0013] S11, acquiring images of bathroom towel rack plugs at different angles, and performing denoising, contrast adjustment, and cropping on the images of bathroom towel rack plugs;

[0014] S12, using image segmentation technology to select a target area of the processed bathroom towel rack plug image, and using a dynamic contour model to generate a preliminary target contour;

[0015] S13, based on the boundary information of the target area, using the B-spline curve to establish an internal and external force model for the target contour, and through the control points and the B-spline curve segmentation strategy, obtain a final target contour that is smooth and fits the target boundary;

[0016] S14. Extracting a feature image of the bathroom towel rack plug based on the final target contour and target area information.

[0017] Furthermore, based on the regional analysis algorithm, the characteristic image of the bathroom towel rack plug was analyzed and the defective areas of the bathroom towel rack plug were identified as follows:

[0018] S21, initializing the parameters and maximum number of iterations of the region analysis algorithm, and generating an initial image feature region set from the bathroom towel rack plug feature image;

[0019] S22, calculate the fitness of each image feature region, find the image feature region with the largest fitness as the current optimal image feature region solution, and compare it with the historical optimal solution. If the current solution is the best, update the historical optimal solution;

[0020] S23, using a ranking selection algorithm to select the best performing image feature region solution from the current image feature region as a parent generation, and generate a new generation of image feature region solutions;

[0021] S24. Based on the conversion probability, choose to perform global or local image feature region update. If global image feature region update is selected, global jumps are simulated by using the Levy step size to increase the diversity of the solution space. If local image feature region update is selected, feature information from the historical optimal solution is introduced to refine the boundary of the feature region.

[0022] S25. When the maximum number of iterations is reached, the region analysis process is terminated, and the optimal image feature region solution is output as the defective region of the bathroom towel rack plug.

[0023] Furthermore, a ranking selection algorithm is used to select the best performing image feature region solution from the current image feature region as the parent, and a new generation of image feature region solutions is generated, including:

[0024] S231. Randomly generate a number of image feature region solutions in the feature dimension space of the image feature region as an initial population;

[0025] S232. Calculate the fitness of each image feature region based on its recognition effect, generate several sub-region solutions for the image feature region with the highest fitness, and allocate the number of offspring using a linear formula;

[0026] S233, the sub-region solution is generated by normal distribution diffusion in the feature dimension space with the parent generation as the center, and similar image feature regions are classified into the same niche, and local diffusion is performed within each niche;

[0027] S234, merging the parent region solution and the child region solution. When the total number exceeds a preset threshold, sorting and screening are performed based on fitness, and the best performing image feature region is retained;

[0028] S235, repeat the fitness evaluation, diffusion generation and region screening process until the maximum number of iterations is reached, output the best performing image feature region as the parent generation, and generate a new generation of image feature region solutions.

[0029] Furthermore, the sub-region solution is generated by normal distribution diffusion in the feature dimension space with the parent as the center, and similar image feature regions are classified into the same niche, and local diffusion is performed within each niche, including:

[0030] S2331. Sort the image feature region solutions in descending order according to fitness. If the population size exceeds a preset threshold, retain several image feature region solutions before the preset threshold.

[0031] S2332: Determine the image feature region with the highest fitness as the center of the first microhabitat, as the core region of the microhabitat;

[0032] S2333. Calculate the Euclidean distance between the remaining image feature region solutions in the population and the center of the current microhabitat. If the distance is less than a preset radius, the solution is included in the microhabitat.

[0033] S2334. Select the one with the highest fitness from the unclassified image feature region solutions as the new microhabitat center, continue to judge and classify them into microhabitats, repeat the division process until all image feature region solutions are classified into corresponding microhabitats, complete the microhabitat classification, and perform local diffusion within each microhabitat.

[0034] Furthermore, the defect classification algorithm is used to identify the defect area, and the defect types of the bathroom towel rack plug are as follows:

[0035] S31, constructing a feature image model of a bathroom towel rack plug, and performing image topology simplification processing based on the structural characteristics of the bathroom towel rack plug;

[0036] S32. Establish a defect segmentation algorithm model to identify independent defect regions of different scales in the feature image of the bathroom towel rack plug;

[0037] S33, setting a minimum defect intensity threshold, retaining defect areas greater than or equal to the threshold, and treating the rest as isolated pixel nodes;

[0038] S34, merging overlapping or connected adjacent defect regions to form a defect partition with the largest range;

[0039] S35. For isolated pixel nodes, calculate the connection relationship between them and each defect area, and divide them based on the minimum feature difference as the criterion;

[0040] S36 , repeatedly calculating and updating the attribution relationship until all image feature points are classified into corresponding defect areas and the defect type is identified.

[0041] Furthermore, a defect segmentation algorithm model is established to identify independent defect regions of different scales in the feature image of the bathroom towel rack plug, including:

[0042] S321, obtaining pixel structure features of a feature image of a blocked head of a bathroom towel rack, and constructing an image topology structure diagram based on an adjacency relationship;

[0043] S322. Using the connection strength between nodes, identify all initial defect connection segments in the bathroom towel rack plug feature image to form a potential defect subgraph;

[0044] S323, constructing a defect partitioning algorithm model, and performing faction stratification on the defect subgraph to identify independent defect area sets;

[0045] S324. Determine the spatial scale of the defective region based on the area occupied by each faction in the bathroom towel rack plug feature image and the connectivity features, and identify a set of independent defective regions of different scales in the bathroom towel rack plug feature image.

[0046] Furthermore, based on the defect type, a scoring algorithm is used to automatically score the quality of bathroom towel rack plugs and generate a defect report. Based on the defect report, the production process is adjusted in real time to optimize the production process to reduce the defective product rate, including:

[0047] S41. Based on the defect type, extract the image features of each defect area and use them as scoring input parameters to evaluate the degree of its impact on the overall quality;

[0048] S42. Apply scoring rules to perform preliminary scoring on each defect area based on defect type and severity, and determine the consistency between the image features and the scoring criteria;

[0049] S43. Conduct multiple rounds of comparison and consistency checks on the preliminary scoring results. If there are scoring deviations, adjust them according to the standards. Finally, confirm the scoring results and generate a defect scoring report.

[0050] S44. Based on the distribution characteristics of various defects in the scoring report, track the changes in related production process parameters and adjust key production process operations to reduce the defective product rate.

[0051] Furthermore, based on the defect type and severity, scoring rules are applied to perform preliminary scoring on each defect area, and the consistency between the image features and the scoring criteria is determined, including:

[0052] S421, setting an initial scoring value for each detected defect area in the image, setting its defect type to an intermediate value by default, and limiting the scoring range;

[0053] S422. Assign a score for the current round based on the image feature location and image features of each defect area according to the scoring rules, and perform weighted calculation based on historical inspection performance to obtain a preliminary score.

[0054] S423. The consistency between the current scoring result and the standard scoring range is judged. If the scoring deviation exceeds the threshold, correction is performed based on the image features to ensure the rationality of the scoring.

[0055] According to another aspect of the present invention, a bathroom towel rack blockage detection system based on AI visual information is provided, the system comprising:

[0056] An image acquisition module is used to acquire images of bathroom towel rack plugs at different angles and extract feature images of bathroom towel rack plugs;

[0057] The regional analysis module is used to analyze the characteristic image of the bathroom towel rack plug based on the regional analysis algorithm and identify the defective area of the bathroom towel rack plug;

[0058] The defect recognition module is used to identify the defect area using the defect segmentation algorithm to obtain the defect type of the bathroom towel rack plug;

[0059] The quality evaluation and optimization module is used to automatically score the quality of bathroom towel rack plugs based on defect types and use a scoring algorithm to generate defect reports. Based on the defect reports, the production process is adjusted in real time to optimize the production process to reduce the defective product rate.

[0060] The beneficial effects of the present invention are:

[0061] 1. The present invention ensures comprehensive coverage of defect information through multi-angle image acquisition and feature extraction. Combined with regional analysis and defect classification algorithms, it automatically identifies defect areas and types to avoid omissions and misjudgments during detection. The defect classification algorithm is used to further distinguish defect types, which helps to accurately evaluate product quality. In addition, through automatic scoring and generation of defect reports, it not only provides real-time quality feedback, but also optimizes production processes based on data, thereby improving production efficiency and product consistency and reducing defective product rates.

[0062] 2. The present invention achieves high-precision defect area identification through a ranking selection algorithm. The algorithm gradually selects the optimal image feature area through fitness calculation and iterative optimization, thereby enhancing the ability to identify complex defects. Moreover, through normal distribution diffusion and niche classification, the algorithm can efficiently handle the diversity of image feature areas, avoiding information omissions and misjudgments. Ultimately, the optimized defect area identification provides an accurate basis for subsequent scoring and quality control, effectively improving the automation and intelligence level of product quality inspection.

[0063] 3. The present invention uses a defect division algorithm to perform topological modeling and regional stratification on the plug image, achieving accurate identification of independent defect areas of different scales. By utilizing the connection relationship and feature difference judgment mechanism, it effectively improves the ability to distinguish between subtle defects and complex defects, thereby enhancing the structural understanding of image processing and the accuracy of defect attribution, and providing a basis for subsequent defect type determination.

[0064] 4. The present invention uses a defect scoring algorithm to achieve a quantitative assessment of the type and severity of plugging defects, combines image features with scoring standards for consistency judgment, effectively improves the accuracy and objectivity of the scoring results, generates a defect report based on the scoring results, and feeds it back to the production link to timely adjust process parameters and optimize the production process, thereby improving the product yield. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0066] Figure 1 This is a flow chart of a method for detecting blockages in bathroom towel racks based on AI visual information according to an embodiment of the present invention;

[0067] Figure 2 This is a principle block diagram of a bathroom towel rack blockage detection system based on AI visual information according to an embodiment of the present invention.

[0068] In the picture:

[0069] 1. Image acquisition module; 2. Regional analysis module; 3. Defect recognition module; 4. Quality evaluation and optimization module. DETAILED DESCRIPTION

[0070] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0071] In the description of the present invention, unless otherwise specified, "plurality" means two or more. In addition, the terms "first", "second", "third", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0072] According to an embodiment of the present invention, a method and system for detecting blockages in bathroom towel racks based on AI visual information are provided.

[0073] The present invention will now be further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, a method for detecting blockage of a bathroom towel rack based on AI visual information according to an embodiment of the present invention includes:

[0074] S1. Acquire images of bathroom towel rack plugs at different angles and extract feature images of bathroom towel rack plugs;

[0075] Specifically, by fixing the plug on a rotating table, the motor controls the gradual rotation, and the CCD camera is fixed to shoot, and a picture is taken every time it rotates a certain angle to obtain images of the bathroom towel rack plug at different angles.

[0076] Specifically, the bathroom towel rack plug feature image includes edge feature image, texture feature image, color feature image, geometric structure feature image, illumination / reflection feature image, etc.

[0077] S2. Analyze the characteristic image of the bathroom towel rack plug based on the regional analysis algorithm to identify the defective area of the bathroom towel rack plug;

[0078] Specifically, the defect area refers to the local area in the characteristic image of the bathroom towel rack plug that is significantly different from the normal structure or surface morphology, including abnormal surface texture areas, structural discontinuity areas, material defect areas, coating defect areas, assembly defect areas, etc.

[0079] S3. Using a defect segmentation algorithm to identify the defect area, and obtaining the defect type of the bathroom towel rack plug;

[0080] Specifically, the defect types of bathroom towel rack plugs include: size defects, height step defects, shape contour defects, round hole eccentricity defects, lower end face defects, chamfer size defects, R angle defects, and tooth rot defects.

[0081] It is important to note that four CCD visual inspection stations are set up to capture images of the plug from different angles. The first station uses a bi-telecentric lens and backlight to detect size, contour, hole eccentricity, and rounded corner defects. The second station uses a 360-degree endoscope to identify tooth decay defects. The third station uses side-viewing to obtain height step dimensions. The fourth station uses bottom-up photography to detect end face defects and chamfers. Combined with a defect classification algorithm, it can identify defect types such as size, height, contour, hole, end face, chamfer, rounded corner, and tooth decay.

[0082] S4. Based on the defect type, the scoring algorithm is used to automatically score the quality of the bathroom towel rack plugs and generate a defect report. Based on the defect report, the production process is adjusted in real time and the production process is optimized to reduce the defective product rate.

[0083] In this optional embodiment, acquiring images of the bathroom towel rack plug at different angles and extracting characteristic images of the bathroom towel rack plug include:

[0084] S11, acquiring images of bathroom towel rack plugs at different angles, and performing denoising, contrast adjustment, and cropping on the images of bathroom towel rack plugs;

[0085] S12, using image segmentation technology to select a target area of the processed bathroom towel rack plug image, and using a dynamic contour model to generate a preliminary target contour;

[0086] S13, based on the boundary information of the target area, using the B-spline curve to establish an internal and external force model for the target contour, and through the control points and the B-spline curve segmentation strategy, obtain a final target contour that is smooth and fits the target boundary;

[0087] S14. Extracting a feature image of the bathroom towel rack plug based on the final target contour and target area information.

[0088] Specifically, the snake algorithm is used to extract the characteristic image of the bathroom towel rack plug, and the contour curve is moved and deformed under the interaction of the internal force of the B-spline curve itself and the external constraint force brought by the bathroom towel rack plug image data. Then, the spatial change mode is determined according to the current position and shape of the contour, and finally the foreground target contour of the bathroom towel rack plug image (that is, the bathroom towel rack plug characteristic image) is approached, thereby achieving accurate extraction of the image foreground and effective improvement of the contour completeness.

[0089] In this optional embodiment, based on the regional analysis algorithm, the characteristic image of the bathroom towel rack plug is analyzed to identify defective areas of the bathroom towel rack plug, including:

[0090] S21, initializing the parameters and maximum number of iterations of the region analysis algorithm, and generating an initial image feature region set from the bathroom towel rack plug feature image;

[0091] S22, calculate the fitness of each image feature region, find the image feature region with the largest fitness as the current optimal image feature region solution, and compare it with the historical optimal solution. If the current solution is the best, update the historical optimal solution;

[0092] S23, using a ranking selection algorithm to select the best performing image feature region solution from the current image feature region as a parent generation, and generate a new generation of image feature region solutions;

[0093] S24. Based on the conversion probability, choose to perform global or local image feature region update. If global image feature region update is selected, global jumps are simulated by using the Levy step size to increase the diversity of the solution space. If local image feature region update is selected, feature information from the historical optimal solution is introduced to refine the boundary of the feature region.

[0094] Specifically, the formula for updating the global image feature region is:

[0095]

[0096] Where, represents the ath image feature region solution in the tth iteration; g * represents the historical optimal solution; δ represents the global step size control coefficient (controls the amplitude of image region adjustment); L represents the Levy distribution step size (simulates cross-region jumps and enhances search diversity); represents the ath image feature region solution in the t+1th iteration.

[0097] Specifically, the formula for updating the local image feature area is:

[0098]

[0099] Where, represents the ath image feature region solution in the tth iteration, represents the ath image feature region solution in the t+1th iteration; and represents the image feature region solution within two randomly selected neighborhoods; ε represents a uniformly distributed random number in [0, 1] (controlling the fine-tuning amplitude).

[0100] S25. When the maximum number of iterations is reached, the region analysis process is terminated, and the optimal image feature region solution is output as the defective region of the bathroom towel rack plug.

[0101] Specifically, the regional analysis algorithm is the discrete flower pollination algorithm, which simulates the pollination process in nature. It combines local and global update strategies, and through continuous search and optimization, it can finally accurately identify the defective area of the bathroom towel rack plug.

[0102] It needs to be explained that after obtaining the feature image of the bathroom towel rack plug, the regional analysis algorithm parameters are initialized and the initial feature region set is generated; the optimal region is selected through fitness calculation, compared with the historical solution and iteratively optimized; the ranking selection algorithm is used to select excellent feature regions to generate a new solution; global or local updates are performed according to the conversion probability, and the search diversity is improved with the help of the Levy step size, and finally the optimal image region is output and marked as the defect area, thereby achieving efficient extraction and precise positioning of the defect area, providing a reliable basis for image detection.

[0103] In this optional embodiment, the best performing image feature region solution is selected from the current image feature regions using a ranking selection algorithm as a parent, and a new generation of image feature region solutions is generated, including:

[0104] S231. Randomly generate a number of image feature region solutions in the feature dimension space of the image feature region as an initial population;

[0105] S232. Calculate the fitness of each image feature region based on its recognition effect, generate several sub-region solutions for the image feature region with the highest fitness, and allocate the number of offspring using a linear formula;

[0106] S233, the sub-region solution is generated by normal distribution diffusion in the feature dimension space with the parent generation as the center, and similar image feature regions are classified into the same niche, and local diffusion is performed within each niche;

[0107] S234, merging the parent region solution and the child region solution. When the total number exceeds a preset threshold, sorting and screening are performed based on fitness, and the best performing image feature region is retained;

[0108] S235, repeat the fitness evaluation, diffusion generation and region screening process until the maximum number of iterations is reached, output the best performing image feature region as the parent generation, and generate a new generation of image feature region solutions.

[0109] Specifically, the ranking selection algorithm is an invasive weed algorithm. This algorithm simulates the weed spread process by generating an initial population, evaluating fitness, performing linear propagation, and performing normal diffusion. Image feature regions with high fitness serve as parents, which then diffuse to generate offspring and assign them to niches. Repeated screening and optimization ultimately yields the optimal image feature region solution.

[0110] It is important to explain that the initial population of image feature region solutions is randomly generated within the feature dimension space, and their recognition performance is calculated to determine their fitness. Regions with high fitness serve as parents, generating several child region solutions. These solutions are then generated by a normal diffusion distribution within the microhabitat. The parent and child generations are merged, and the optimal region is selected. This process is repeated until the iteration limit is reached. Finally, the best performing image feature region is output for subsequent defect recognition, thereby enhancing the ability to explore and converge the solution space and providing a high-quality initial feature region foundation for subsequent defect recognition.

[0111] In this optional embodiment, the subregion solution is generated by performing normal distribution diffusion in the feature dimension space with the parent as the center, classifying similar image feature regions into the same niche, and performing local diffusion within each niche, including:

[0112] S2331. Sort the image feature region solutions in descending order according to fitness. If the population size exceeds a preset threshold, retain several image feature region solutions before the preset threshold.

[0113] S2332: Determine the image feature region with the highest fitness as the center of the first microhabitat, as the core region of the microhabitat;

[0114] S2333. Calculate the Euclidean distance between the remaining image feature region solutions in the population and the center of the current microhabitat. If the distance is less than a preset radius, the solution is included in the microhabitat.

[0115] S2334. Select the one with the highest fitness from the unclassified image feature region solutions as the new microhabitat center, continue to judge and classify them into microhabitats, repeat the division process until all image feature region solutions are classified into corresponding microhabitats, complete the microhabitat classification, and perform local diffusion within each microhabitat.

[0116] It should be explained that the image feature area solutions are first sorted in descending order according to fitness, and the first few solutions are retained. The solution with the highest fitness is used as the center of the first microhabitat, and then the distance between other solutions and the center is calculated. If the distance is less than the preset radius, it is included in the microhabitat. Next, the solution with the highest fitness is selected from the unclassified solutions as the center of the new microhabitat, and classification is continued and this process is repeated until all solutions are classified. Finally, local diffusion is performed within each microhabitat to optimize the image feature area solutions, thereby achieving fine division and optimization of the feature areas, improving the stability of image analysis and local search capabilities, and providing a better solution space for defect identification.

[0117] In this optional embodiment, the defect area is identified using a defect classification algorithm, and the defect types of the bathroom towel rack plug include:

[0118] S31, constructing a feature image model of a bathroom towel rack plug, and performing image topology simplification processing based on the structural characteristics of the bathroom towel rack plug;

[0119] S32. Establish a defect segmentation algorithm model to identify independent defect regions of different scales in the feature image of the bathroom towel rack plug;

[0120] S33, setting a minimum defect intensity threshold, retaining defect areas greater than or equal to the threshold, and treating the rest as isolated pixel nodes;

[0121] S34, merging overlapping or connected adjacent defect regions to form a defect partition with the largest range;

[0122] S35. For isolated pixel nodes, calculate the connection relationship between them and each defect area, and divide them based on the minimum feature difference as the criterion;

[0123] S36 , repeatedly calculating and updating the attribution relationship until all image feature points are classified into corresponding defect areas and the defect type is identified.

[0124] Specifically, the defect segmentation algorithm is a clique filtering algorithm. It identifies defective areas in bathroom towel rack plugs by constructing an image model, simplifying topology, and segmenting defects. First, the algorithm identifies defective areas of varying scales and sets an intensity threshold to filter out isolated pixels. It then merges adjacent defective areas and calculates the connectivity between isolated pixels and defective areas. Finally, it uses minimum feature difference segmentation to accurately classify all image feature points and identify the specific defect type.

[0125] It should be explained that the plug image is first topologically simplified to extract key structural features. A defect segmentation algorithm is then used to identify multi-scale defect regions, set intensity thresholds, and remove isolated pixels. Connected or overlapping defect regions are merged to form complete partitions, and isolated pixels are assigned to corresponding defect regions based on the minimum feature difference. Finally, the attribution relationship is repeatedly updated until all image feature points are classified and the specific defect type of the plug is accurately identified. This allows for precise segmentation and identification of various defect regions, effectively improving the reliability and accuracy of plug quality inspection.

[0126] In this optional embodiment, a defect segmentation algorithm model is established to identify a set of independent defect regions of different scales in a feature image of a bathroom towel rack plug, including:

[0127] S321, obtaining pixel structure features of a feature image of a blocked head of a bathroom towel rack, and constructing an image topology structure diagram based on an adjacency relationship;

[0128] S322. Using the connection strength between nodes, identify all initial defect connection segments in the bathroom towel rack plug feature image to form a potential defect subgraph;

[0129] S323, constructing a defect partitioning algorithm model, and performing faction stratification on the defect subgraph to identify independent defect area sets;

[0130] S324. Determine the spatial scale of the defective region based on the area occupied by each faction in the bathroom towel rack plug feature image and the connectivity features, and identify a set of independent defective regions of different scales in the bathroom towel rack plug feature image.

[0131] It is important to explain that the pixel structure information of the plug feature image is first extracted, and an adjacency graph is constructed to form the image topology. Subsequently, potential defect connected segments are extracted based on the node connection strength, and a defect subgraph is constructed. The subgraph is divided into faction hierarchies to identify structurally independent defect regions. Finally, the spatial scale of each faction is determined by combining its area and connectivity characteristics in the image, completing the identification and classification of defect regions of different scales. This allows for efficient identification and classification of defect regions of different scales, improving the precision and accuracy of defect detection.

[0132] In this optional embodiment, a scoring algorithm is used to automatically score the quality of bathroom towel rack plugs based on the defect type and generate a defect report. Based on the defect report, the production process is adjusted in real time to optimize the production process to reduce the defective product rate, including:

[0133] S41. Based on the defect type, extract the image features of each defect area and use them as scoring input parameters to evaluate the degree of its impact on the overall quality;

[0134] S42. Apply scoring rules to perform preliminary scoring on each defect area based on defect type and severity, and determine the consistency between the image features and the scoring criteria;

[0135] S43. Conduct multiple rounds of comparison and consistency checks on the preliminary scoring results. If there are scoring deviations, adjust them according to the standards. Finally, confirm the scoring results and generate a defect scoring report.

[0136] S44. Based on the distribution characteristics of various defects in the scoring report, track the changes in related production process parameters and adjust key production process operations to reduce the defective product rate.

[0137] Specifically, the scoring algorithm uses a practical Byzantine fault-tolerance algorithm. Through multiple rounds of scoring comparisons and consistency checks, the algorithm eliminates abnormal or deviant scoring results, ensuring the reliability of each defect scoring data. The algorithm reaches consensus among scoring nodes, preventing individual misjudgments from affecting overall scoring accuracy. Ultimately, it generates a reliable defect scoring report, providing a stable basis for production process optimization.

[0138] It is important to explain that the system first extracts image features from various defect areas and assigns a preliminary score based on defect type and severity. Multiple rounds of scoring consistency checks are then conducted to correct any discrepancies and generate a final defect score report. Based on the defect distribution in this report, the system then traces the relevant process steps, identifies parameter fluctuations that lead to defects, and adjusts production operation strategies in real time, thereby optimizing the process flow, improving overall plug quality, and reducing defective product rates.

[0139] In this optional embodiment, based on the defect type and severity, a scoring rule is applied to perform a preliminary score on each defect area, and the consistency between the image features and the scoring criteria is determined, including:

[0140] S421, setting an initial scoring value for each detected defect area in the image, setting its defect type to an intermediate value by default, and limiting the scoring range;

[0141] S422. Assign a score for the current round based on the image feature location and image features of each defect area according to the scoring rules, and perform weighted calculation based on historical inspection performance to obtain a preliminary score.

[0142] S423. The consistency between the current scoring result and the standard scoring range is judged. If the scoring deviation exceeds the threshold, correction is performed based on the image features to ensure the rationality of the scoring.

[0143] It is important to explain that an initial score is first set for each detected defect area, with the defect type defaulted to a medium level. Scoring is then performed based on image features such as location and shape, combined with scoring rules, and weighted based on historical inspection results to form a preliminary score. Finally, the preliminary score is compared with the standard scoring range for consistency. Any deviations are corrected based on the image features. This ensures the accuracy and consistency of defect scoring, provides a reliable basis for subsequent quality assessment and production optimization, and ensures the accuracy and rationality of the scoring results.

[0144] According to another embodiment of the present invention, Figure 2 As shown, a bathroom towel rack blockage detection system based on AI visual information is also provided, which includes:

[0145] Image acquisition module 1, used to acquire images of bathroom towel rack plugs at different angles and extract feature images of bathroom towel rack plugs;

[0146] The regional analysis module 2 is used to analyze the characteristic image of the bathroom towel rack plug based on the regional analysis algorithm to identify the defective area of the bathroom towel rack plug;

[0147] Defect identification module 3 is used to identify the defect area using a defect classification algorithm to obtain the defect type of the bathroom towel rack plug;

[0148] The quality evaluation and optimization module 4 is used to automatically score the quality of bathroom towel rack plugs based on the defect type and use a scoring algorithm to generate a defect report; and based on the defect report, it adjusts the production process in real time and optimizes the production process to reduce the defective product rate.

[0149] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A bathroom towel rack plug detection method based on AI visual information, characterized in that: include: S1. Acquire images of bathroom towel rack plugs at different angles and extract feature images of bathroom towel rack plugs; S2. Analyze the characteristic image of the bathroom towel rack plug based on the regional analysis algorithm to identify the defective area of the bathroom towel rack plug; S3. Using a defect segmentation algorithm to identify the defect area, and obtaining the defect type of the bathroom towel rack plug; S4. Based on the defect type, the scoring algorithm is used to automatically score the quality of the bathroom towel rack plug and generate a defect report; Based on defect reports, the production process is adjusted in real time and the production process is optimized to reduce the defective product rate.

2. The method for detecting blockages in bathroom towel racks based on AI visual information according to claim 1, characterized in that: The method of obtaining images of the bathroom towel rack plug at different angles and extracting characteristic images of the bathroom towel rack plug includes: S11, acquiring images of bathroom towel rack plugs at different angles, and performing denoising, contrast adjustment, and cropping on the images of bathroom towel rack plugs; S12, using image segmentation technology to select a target area of the processed bathroom towel rack plug image, and using a dynamic contour model to generate a preliminary target contour; S13, based on the boundary information of the target area, using the B-spline curve to establish an internal and external force model for the target contour, and through the control points and the B-spline curve segmentation strategy, obtain a final target contour that is smooth and fits the target boundary; S14. Extracting a feature image of the bathroom towel rack plug based on the final target contour and target area information.

3. The method for detecting blockages in bathroom towel racks based on AI visual information according to claim 1, characterized in that: The regional analysis algorithm is used to analyze the characteristic image of the bathroom towel rack plug and identify defective areas of the bathroom towel rack plug, including: S21, initializing the parameters and maximum number of iterations of the region analysis algorithm, and generating an initial image feature region set from the bathroom towel rack plug feature image; S22, calculate the fitness of each image feature region, find the image feature region with the largest fitness as the current optimal image feature region solution, and compare it with the historical optimal solution. If the current solution is the best, update the historical optimal solution; S23, using a ranking selection algorithm to select the best performing image feature region solution from the current image feature region as a parent generation, and generate a new generation of image feature region solutions; S24. Based on the conversion probability, choose to perform global or local image feature region update. If global image feature region update is selected, global jumps are simulated by using the Levy step size to increase the diversity of the solution space. If local image feature region update is selected, feature information from the historical optimal solution is introduced to refine the boundary of the feature region. S25. When the maximum number of iterations is reached, the region analysis process is terminated, and the optimal image feature region solution is output as the defective region of the bathroom towel rack plug.

4. The method for detecting blockages in bathroom towel racks based on AI visual information according to claim 3 is characterized in that: The method of selecting the best performing image feature region solution from the current image feature region using a ranking selection algorithm as a parent and generating a new generation of image feature region solutions includes: S231. Randomly generate a number of image feature region solutions in the feature dimension space of the image feature region as an initial population; S232. Calculate the fitness of each image feature region based on its recognition effect, generate several sub-region solutions for the image feature region with the highest fitness, and allocate the number of offspring using a linear formula; S233, the sub-region solution is generated by normal distribution diffusion in the feature dimension space with the parent generation as the center, and similar image feature regions are classified into the same niche, and local diffusion is performed within each niche; S234, merging the parent region solution and the child region solution. When the total number exceeds a preset threshold, sorting and screening are performed based on fitness, and the best performing image feature region is retained; S235, repeat the fitness evaluation, diffusion generation and region screening process until the maximum number of iterations is reached, output the best performing image feature region as the parent generation, and generate a new generation of image feature region solutions.

5. The method for detecting blockages in bathroom towel racks based on AI visual information according to claim 4 is characterized in that: The sub-region solution is centered on the parent generation and is generated by normal distribution diffusion in the feature dimension space. Similar image feature regions are classified into the same niche, and local diffusion is performed within each niche, including: S2331. Sort the image feature region solutions in descending order according to fitness. If the population size exceeds a preset threshold, retain several image feature region solutions before the preset threshold. S2332: Determine the image feature region with the highest fitness as the center of the first microhabitat, as the core region of the microhabitat; S2333. Calculate the Euclidean distance between the remaining image feature region solutions in the population and the center of the current microhabitat. If the distance is less than a preset radius, the solution is included in the microhabitat. S2334. Select the one with the highest fitness from the unclassified image feature region solutions as the new microhabitat center, continue to judge and classify them into microhabitats, repeat the division process until all image feature region solutions are classified into corresponding microhabitats, complete the microhabitat classification, and perform local diffusion within each microhabitat.

6. The method for detecting blockages in bathroom towel racks based on AI visual information according to claim 1, characterized in that: The defect area is identified by using the defect division algorithm, and the defect types of the bathroom towel rack plug include: S31, constructing a feature image model of a bathroom towel rack plug, and performing image topology simplification processing based on the structural characteristics of the bathroom towel rack plug; S32. Establish a defect segmentation algorithm model to identify independent defect regions of different scales in the feature image of the bathroom towel rack plug; S33, setting a minimum defect intensity threshold, retaining defect areas greater than or equal to the threshold, and treating the rest as isolated pixel nodes; S34, merging overlapping or connected adjacent defect regions to form a defect partition with the largest range; S35. For isolated pixel nodes, calculate the connection relationship between them and each defect area, and divide them based on the minimum feature difference as the criterion; S36 , repeatedly calculating and updating the attribution relationship until all image feature points are classified into corresponding defect areas and the defect type is identified.

7. The method for detecting blockages in bathroom towel racks based on AI visual information according to claim 6, characterized in that: The defect segmentation algorithm model is established to identify a set of independent defect regions of different scales in the feature image of the bathroom towel rack plug, including: S321, obtaining pixel structure features of a feature image of a blocked head of a bathroom towel rack, and constructing an image topology structure diagram based on an adjacency relationship; S322. Using the connection strength between nodes, identify all initial defect connection segments in the bathroom towel rack plug feature image to form a potential defect subgraph; S323, constructing a defect partitioning algorithm model, and performing faction stratification on the defect subgraph to identify independent defect area sets; S324. Determine the spatial scale of the defective region based on the area occupied by each faction in the bathroom towel rack plug feature image and the connectivity features, and identify a set of independent defective regions of different scales in the bathroom towel rack plug feature image.

8. The method for detecting blockages in bathroom towel racks based on AI visual information according to claim 1, characterized in that: The defect type is combined with a scoring algorithm to automatically score the quality of the bathroom towel rack plug and generate a defect report; Based on defect reports, production processes are adjusted in real time to optimize production processes to reduce defective product rates, including: S41. Based on the defect type, extract the image features of each defect area and use them as scoring input parameters to evaluate the degree of its impact on the overall quality; S42. Apply scoring rules to perform preliminary scoring on each defect area based on defect type and severity, and determine the consistency between the image features and the scoring criteria; S43. Conduct multiple rounds of comparison and consistency checks on the preliminary scoring results. If there are scoring deviations, adjust them according to the standards. Finally, confirm the scoring results and generate a defect scoring report. S44. Based on the distribution characteristics of various defects in the scoring report, track the changes in related production process parameters and adjust key production process operations to reduce the defective product rate.

9. The method for detecting blockages in bathroom towel racks based on AI visual information according to claim 8, characterized in that: The scoring rules are applied to perform preliminary scoring on each defect area based on the defect type and severity, and the consistency between the image features and the scoring criteria is determined, including: S421, setting an initial scoring value for each detected defect area in the image, setting its defect type to an intermediate value by default, and limiting the scoring range; S422. Assign a score for the current round based on the image feature location and image features of each defect area according to the scoring rules, and perform weighted calculation based on historical inspection performance to obtain a preliminary score. S423. The consistency between the current scoring result and the standard scoring range is judged. If the scoring deviation exceeds the threshold, correction is performed based on the image features to ensure the rationality of the scoring.

10. A bathroom towel rack blockage detection system based on AI visual information, used to implement the bathroom towel rack blockage detection method based on AI visual information according to any one of claims 1 to 9, characterized in that: The system includes: An image acquisition module is used to acquire images of bathroom towel rack plugs at different angles and extract feature images of bathroom towel rack plugs; The regional analysis module is used to analyze the characteristic image of the bathroom towel rack plug based on the regional analysis algorithm and identify the defective area of the bathroom towel rack plug; The defect recognition module is used to identify the defect area using the defect segmentation algorithm to obtain the defect type of the bathroom towel rack plug; The quality evaluation and optimization module is used to automatically score the quality of bathroom towel rack plugs based on defect types and use a scoring algorithm to generate defect reports. Based on the defect reports, the production process is adjusted in real time to optimize the production process to reduce the defective product rate.

Images