Seat surface defect detection method and device

Abstract

The invention provides a seat surface defect detection method and device, and belongs to the technical field of surface quality detection. A seat surface defect detection method comprises the following steps: controlling a tactile scanning execution mechanism to move along a preset detection track of a seat surface, performing contact scanning on the seat surface, and obtaining contact pressure data, displacement feedback data and surface rebound data of a corresponding position; according to the contact pressure data, the displacement feedback data and the surface rebound data, extracting a surface flatness feature, a hardness anomaly feature and a local deformation recovery feature of the surface of the seat, and determining a tactile detection anomaly area according to the surface flatness feature, the hardness anomaly feature and the local deformation recovery feature; and performing visual image acquisition on an abnormal region obtained by touch detection, extracting texture abnormal features and color abnormal features of the corresponding region, and generating a visual review result.

Description

Technical Field

[0001] This invention relates to the field of surface quality inspection technology, and in particular to a method and apparatus for detecting surface defects in seats. Background Technology

[0002] During the production, assembly, and final inspection of seats, defects such as bulges, depressions, foreign object inclusions, surface wear, and uneven stitching frequently appear on the seat surface. Some of these defects manifest as abnormal surface geometry, others as abnormal internal support, and still others as abnormal changes in surface texture and color. A single detection method is often insufficient to reliably identify multiple defects.

[0003] In existing technologies, surface defect detection for seats mostly relies on manual visual inspection or single-vision inspection schemes. While manual visual inspection can identify obvious appearance defects, its efficiency is greatly affected by human experience, resulting in poor consistency and insufficient ability to identify minor bulges, localized collapses, and uneven hardness. Although single-vision inspection schemes can identify issues such as abnormal surface color, missing texture, and seam misalignment through image analysis, they often struggle to accurately diagnose bulges, collapses, or support abnormalities caused by intact surface covering material but abnormal internal filling conditions, relying solely on two-dimensional image information.

[0004] On the other hand, some defects are not obvious on the surface, but will show abnormal rebound, abnormal support differences, or abnormal local undulations after contact and pressure. For example, local accumulation of padding material inside the seat may cause surface bulges, which may only appear as slight contour changes visually, but abnormal contact pressure and abnormal local deformation recovery are more easily detected through tactile contact scanning. As another example, collapse defects caused by local voids or insufficient filling may be obscured by the surface tension in static images, but will show obvious differences in displacement feedback during pressure.

[0005] Furthermore, while single-sensory detection schemes can acquire contact pressure data, displacement feedback data, and surface rebound data, they still suffer from insufficient ability to distinguish defect types that rely on surface texture and color, such as foreign object inclusions, surface wear, and uneven seams. Especially when different defects exhibit local similarities in tactile characteristics, relying solely on contact response results can easily lead to misclassification of defect categories. For example, some uneven seam areas may also be accompanied by local undulations, and some foreign object inclusion areas may exhibit abnormal softness or hardness; without visual image acquisition and verification of texture and color features, accurate classification is difficult. Summary of the Invention

[0006] The purpose of this invention is to provide a method and apparatus for detecting surface defects of seats, so as to solve the problems in the prior art that it is difficult to identify internal support abnormalities by single visual inspection, difficult to distinguish surface texture and color abnormalities by single tactile inspection, and insufficient accuracy in classifying various seat surface defects.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A method for detecting defects on the surface of a seat includes the following steps:

[0009] The tactile scanning actuator is controlled to move along a preset detection trajectory on the seat surface to perform contact scanning on the seat surface and acquire contact pressure data, displacement feedback data and surface rebound data at the corresponding positions.

[0010] Based on the contact pressure data, displacement feedback data, and surface rebound data, the surface flatness features, softness / hardness anomaly features, and local deformation recovery features of the seat surface are extracted, and the tactile detection abnormal area is determined based on the surface flatness features, softness / hardness anomaly features, and local deformation recovery features.

[0011] Visual images are acquired from the abnormal areas obtained by tactile detection, and the texture and color abnormal features of the corresponding areas are extracted to generate visual verification results.

[0012] Based on the tactile detection results and visual verification results, the defects of the seat surface, such as bulges, depressions, foreign object inclusions, surface wear, and uneven stitching, are comprehensively judged, and the corresponding defect categories and defect locations are output.

[0013] Furthermore, the control tactile scanning actuator moves along a preset detection trajectory on the seat surface to perform contact scanning on the seat surface and acquire contact pressure data, displacement feedback data, and surface rebound data at the corresponding positions, including the following steps:

[0014] Obtain the contour boundary information, surface undulation information, and partition position information of the seat surface, and generate the preset detection trajectory based on the contour boundary information, surface undulation information, and partition position information;

[0015] The corresponding scanning step distance, contact depth, and moving speed are set according to the curvature and boundary changes of different areas of the seat surface;

[0016] The tactile scanning actuator is controlled to move point by point along the preset detection trajectory, and a preset contact force is applied to the seat surface at each scanning position;

[0017] During the movement of the tactile scanning actuator along the preset detection trajectory, contact pressure data, displacement feedback data, and surface rebound data corresponding to each scanning position are collected simultaneously, and the correspondence between the contact pressure data, displacement feedback data, and surface rebound data and the scanning position is established.

[0018] Furthermore, based on the contact pressure data, displacement feedback data, and surface rebound data, surface flatness features, softness / hardness anomaly features, and local deformation recovery features of the seat surface are extracted. Based on these surface flatness features, softness / hardness anomaly features, and local deformation recovery features, tactile detection abnormal areas are determined, including the following steps:

[0019] The pressure fluctuation and contact undulation of the seat surface at different scanning positions are extracted based on the contact pressure data to generate the surface flatness feature; the displacement change and support difference of the seat surface at different scanning positions are extracted based on the displacement feedback data to generate the softness and hardness anomaly feature.

[0020] Based on the surface rebound data, the recovery time, recovery amplitude, and recovery stability of the seat surface after being compressed are extracted to generate the local deformation recovery characteristics.

[0021] By jointly analyzing the surface smoothness features, softness and hardness anomaly features, and local deformation recovery features, tactile anomaly scores corresponding to different scanning positions are obtained, and the tactile detection anomaly areas are determined based on the tactile anomaly scores.

[0022] Furthermore, the methods for determining abnormal tactile detection areas based on the surface smoothness characteristics, softness / hardness anomaly characteristics, and local deformation recovery characteristics include:

[0023] The abnormal sections of continuous surface undulations are determined based on the changes in surface smoothness characteristics between adjacent scanning positions.

[0024] Local support anomaly segments are determined based on the changes in soft and hard anomaly characteristics between adjacent scan locations;

[0025] Locally recovered abnormal segments are determined based on the changes in local deformation recovery characteristics between adjacent scan locations;

[0026] The abnormal sections of continuous surface undulation, local support, and local recovery are spatially superimposed to obtain candidate tactile abnormality sections. The tactile detection abnormality region is then determined based on the continuous length, abnormal intensity, and location clustering relationship of the candidate tactile abnormality sections.

[0027] Furthermore, the step of acquiring visual images of the abnormal regions obtained by tactile detection, extracting the texture and color anomaly features of the corresponding regions, and generating visual verification results includes the following steps:

[0028] The visual image acquisition center, acquisition angle, and acquisition distance are determined based on the range information of the abnormal tactile detection area.

[0029] The visual acquisition device is controlled to perform directional visual image acquisition on the abnormal tactile detection area to obtain the corresponding local area image;

[0030] Based on the local region image, extract texture anomaly features that characterize the continuous state of surface texture, and color anomaly features that characterize the state of surface color change.

[0031] The texture anomaly features and color anomaly features are mapped to the spatial locations of the tactile detection anomaly areas to generate the visual verification result.

[0032] Furthermore, the method for generating the visual verification result includes:

[0033] Based on the texture anomaly features, calculate the degree of texture continuity change, texture breakage, and texture misalignment within the tactile detection anomaly area;

[0034] Based on the aforementioned color anomaly characteristics, calculate the degree of color shift, the degree of uneven color distribution, and the degree of color abrupt change within the abnormal tactile detection area;

[0035] The degree of continuous variation in texture, degree of texture breakage, degree of texture misalignment, degree of color shift, degree of uneven color distribution, and degree of color mutation are combined to obtain the visual anomaly score.

[0036] Based on the comparison between the visual anomaly score and the preset visual verification threshold, a visual verification result corresponding to the tactile detection anomaly area is generated.

[0037] Furthermore, based on the tactile detection results and visual verification results, a comprehensive judgment is made on defects such as bulges, depressions, foreign object inclusions, surface wear, and uneven stitching on the seat surface, and the corresponding defect categories and locations are output, including the following steps:

[0038] The surface smoothness features, softness and hardness anomaly features, and local deformation recovery features in the tactile detection results are fused with the texture anomaly features and color anomaly features in the visual verification results to generate defect candidate judgment results.

[0039] Based on the combination relationship of various features in the defect candidate determination results, matching analysis is performed on defects such as bulges, collapses, foreign matter inclusions, surface wear, and uneven stitching.

[0040] When the surface smoothness features are characterized by local bulges and abnormal local deformation recovery features, it is judged as a bulge; when the surface smoothness features are characterized by local depressions and abnormal softness / hardness features, it is judged as a collapse.

[0041] When abnormal texture and color features are localized to be different from the surrounding area, they are identified as foreign matter inclusions. When abnormal texture features are characterized by missing or disordered surface texture, they are identified as surface wear. When abnormal texture features are characterized by abnormal continuity in the suture line area, they are identified as uneven suture defects.

[0042] Furthermore, the output method for the defect location includes:

[0043] The initial position of the tactile detection abnormal area is determined based on the scanning position coordinates corresponding to the movement of the tactile scanning actuator along the preset detection trajectory.

[0044] The image position of the visual verification result is determined based on the imaging area coordinates corresponding to the visual image acquisition device when it acquires visual images of the tactile detection abnormal area.

[0045] By performing coordinate mapping and region alignment between the initial position and the image position, the target position of the corresponding defect on the seat surface is obtained;

[0046] The target location is associated with the defect category and output to form the defect location.

[0047] Furthermore, it also includes an anomaly review and update step, which includes:

[0048] When the tactile detection result is inconsistent with the visual verification result, a second tactile scan is performed on the corresponding tactile detection abnormal area.

[0049] Based on the secondary tactile scan, the contact pressure data, displacement feedback data and surface rebound data of the corresponding positions are re-acquired, and the surface flatness features, soft and hard anomaly features and local deformation recovery features are re-extracted.

[0050] Visual images of the tactile detection abnormal areas are re-acquired, and the texture and color abnormal features of the corresponding areas are re-extracted.

[0051] Based on the re-extracted surface flatness features, softness and hardness anomaly features, local deformation recovery features, texture anomaly features, and color anomaly features, the defects such as bulges, collapses, foreign matter inclusions, surface wear, and uneven seams are updated and judged, and the corresponding defect categories and defect locations are re-output.

[0052] A seat surface defect detection device based on robotic tactile scanning and visual verification includes:

[0053] The tactile scanning module controls the tactile scanning actuator to move along a preset detection trajectory on the seat surface, perform contact scanning on the seat surface, and acquire contact pressure data, displacement feedback data, and surface rebound data at the corresponding positions.

[0054] The tactile feature analysis module extracts surface flatness features, softness / hardness anomaly features, and local deformation recovery features of the seat surface based on the contact pressure data, displacement feedback data, and surface rebound data, and determines the tactile detection abnormal area based on the surface flatness features, softness / hardness anomaly features, and local deformation recovery features.

[0055] The visual verification module acquires visual images of abnormal areas obtained by tactile detection, extracts texture and color abnormality features of the corresponding areas, and generates visual verification results.

[0056] The comprehensive judgment module, based on tactile detection results and visual verification results, comprehensively judges defects on the seat surface such as bulges, depressions, foreign object inclusions, surface wear, and uneven stitching, and outputs the corresponding defect category and defect location. Compared with the prior art, the present invention has at least the following beneficial effects:

[0057] First, the present invention first performs a contact scan on the seat surface through a tactile scanning actuator, and then performs directional visual image acquisition on the abnormal tactile detection areas, forming a detection process of "tactile screening first, visual verification later", thereby avoiding high-load full visual fine inspection of the entire seat surface and improving detection efficiency.

[0058] Secondly, this invention simultaneously utilizes contact pressure data, displacement feedback data, and surface rebound data to extract surface flatness features, soft and hard anomaly features, and local deformation recovery features. This enables the identification of internal support anomalies and local rebound anomalies that are not easily visible in visual images, which is beneficial for improving the detection capability of bulge and collapse defects.

[0059] Third, by extracting texture and color abnormality features from abnormal areas of tactile detection, this invention enables visual verification of defects such as foreign matter inclusions, surface wear, and uneven stitching, which helps to improve the accuracy of distinguishing between different defect categories.

[0060] Fourth, this invention integrates the tactile detection results with the visual verification results, and then performs a comprehensive judgment based on different feature combinations, making the judgment logic for defects such as bulges, collapses, foreign matter inclusions, surface wear, and uneven stitching more complete, which helps to reduce misjudgments caused by a single sensing dimension.

[0061] Fifth, by mapping and aligning the scanning position coordinates with the imaging area coordinates, this invention can output the target location associated with the defect category, which is beneficial for subsequent rework, re-inspection and quality traceability.

[0062] Sixth, the present invention sets up an anomaly review and update step, which performs a second tactile scan and re-acquisition of visual images when the tactile detection result is inconsistent with the visual review result, which helps to improve the stability of the determination of boundary anomaly areas and complex anomaly areas. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of the overall process of a method for detecting surface defects in a seat according to the present invention;

[0064] Figure 2 This is a schematic diagram of the module structure of a seat surface defect detection device according to the present invention;

[0065] Figure 3 This is a schematic diagram of the tactile scanning and abnormal area identification process in this invention;

[0066] Figure 4 This is a schematic diagram illustrating the fusion and determination of tactile detection results and visual verification results in this invention. Detailed Implementation

[0067] To make the objectives, technical solutions, and beneficial effects of this invention clearer, the invention will be further described in detail below with reference to specific embodiments. It should be noted that the following specific embodiments are based on the technical solutions described in the claims and are used to illustrate the implementation logic, data flow relationships, feature extraction relationships, verification and judgment relationships, and position output relationships of this invention, rather than limiting the scope of protection of this invention. In some embodiments, the seat surface defect detection method of this invention can be applied to quality inspection scenarios for seat cushions, backrests, headrests, armrests, and other seat components with covered surfaces and internal filling structures.

[0068] like Figure 1-4 As shown, in some embodiments, the overall process of the present invention can be summarized as follows: First, a preset detection trajectory is generated based on the contour boundary information, surface undulation information, and partition position information of the seat surface. The tactile scanning execution mechanism moves along the preset detection trajectory to perform contact scanning on the seat surface, acquiring contact pressure data, displacement feedback data, and surface rebound data. Then, surface flatness features, softness / hardness anomaly features, and local deformation recovery features are extracted based on the contact pressure data, displacement feedback data, and surface rebound data. Tactile detection anomaly areas are determined based on the surface flatness features, softness / hardness anomaly features, and local deformation recovery features. Afterward, visual images are acquired from the tactile detection anomaly areas, and texture anomaly features and color anomaly features of the corresponding areas are extracted to generate visual verification results. Finally, based on the tactile detection results and visual verification results, bulges, collapses, foreign object inclusions, surface wear, and uneven stitching defects are comprehensively judged, and the corresponding defect categories and defect locations are output.

[0069] S100 controls the tactile scanning actuator to move along a preset detection trajectory on the seat surface, performs contact scanning on the seat surface, and acquires contact pressure data, displacement feedback data, and surface rebound data at the corresponding positions.

[0070] S100 specifically includes the following steps:

[0071] Obtain the contour boundary information, surface undulation information, and partition position information of the seat surface, and generate the preset detection trajectory based on the contour boundary information, surface undulation information, and partition position information;

[0072] The corresponding scanning step distance, contact depth, and moving speed are set according to the curvature and boundary changes of different areas of the seat surface;

[0073] The tactile scanning actuator is controlled to move point by point along the preset detection trajectory, and a preset contact force is applied to the seat surface at each scanning position;

[0074] During the movement of the tactile scanning actuator along the preset detection trajectory, contact pressure data, displacement feedback data, and surface rebound data corresponding to each scanning position are collected simultaneously, and the correspondence between the contact pressure data, displacement feedback data, and surface rebound data and the scanning position is established.

[0075] In the above steps, the input content mainly includes the geometric shape information of the seat surface and the motion control command of the tactile scanning actuator; the preset detection trajectory is the basis for the movement path of the tactile scanning actuator on the seat surface. When the tactile scanning actuator moves along the preset detection trajectory, it applies a controllable contact force to the seat surface and acquires the corresponding contact pressure data, displacement feedback data and surface rebound data at each scanning position.

[0076] Among them, contact pressure data is used to characterize the pressure state and local undulation state when the tactile scanning actuator contacts the seat surface, displacement feedback data is used to characterize the displacement response state and local support state of the seat surface under the preset contact force, and surface rebound data is used to characterize the recovery state of the seat surface when the pressure is released. By acquiring the above three types of data simultaneously, the state of the seat surface and its internal support structure can be characterized from three dimensions: pressure, displacement and rebound.

[0077] In some embodiments, the tactile scanning actuator can be mounted on the robot end effector and achieve stable movement along a preset detection trajectory via a linear module, articulated robot, or multi-axis motion platform. The preset contact force can be set according to the seat material type, surface thickness, filling layer hardness, and detection area characteristics to avoid excessive pressure damaging the seat surface and to avoid insufficient pressure making it difficult to detect support differences. The output of this step is contact pressure data, displacement feedback data, and surface rebound data corresponding to each scanning position.

[0078] In one implementation scenario, when inspecting the surface of a seat cushion, a serpentine or partitioned pre-set detection trajectory is first generated according to the cushion boundary and the central pressure area. Then, the tactile scanning execution mechanism sequentially performs contact scanning on the edge area, the central area, and the adjacent area of ​​the seam line, thereby forming an original tactile detection dataset covering the entire seat surface.

[0079] S200: Extract surface flatness features, softness / hardness anomaly features, and local deformation recovery features of the seat surface based on the contact pressure data, displacement feedback data, and surface rebound data, and determine the tactile detection abnormal area based on the surface flatness features, softness / hardness anomaly features, and local deformation recovery features.

[0080] In the above steps, the inputs are the contact pressure data, displacement feedback data, and surface rebound data output by S100. The core of this step is to convert the raw tactile detection data into high-level feature results that can be used for anomaly localization and defect determination, and to further determine the tactile detection anomaly area.

[0081] S200 specifically includes the following steps:

[0082] The pressure fluctuation and contact undulation of the seat surface at different scanning positions are extracted based on the contact pressure data to generate the surface flatness feature.

[0083] The displacement change and support difference of the seat surface at different scanning positions are extracted based on the displacement feedback data to generate the soft and hard anomaly features.

[0084] Based on the surface rebound data, the recovery time, recovery amplitude, and recovery stability of the seat surface after being compressed are extracted to generate the local deformation recovery characteristics.

[0085] By jointly analyzing the surface smoothness features, softness and hardness anomaly features, and local deformation recovery features, tactile anomaly scores corresponding to different scanning positions are obtained, and the tactile detection anomaly areas are determined based on the tactile anomaly scores.

[0086] The methods for determining abnormal tactile detection areas based on the surface smoothness characteristics, softness / hardness anomaly characteristics, and local deformation recovery characteristics include:

[0087] The abnormal sections of continuous surface undulations are determined based on the changes in surface smoothness characteristics between adjacent scanning positions.

[0088] Local support anomaly segments are determined based on the changes in soft and hard anomaly characteristics between adjacent scan locations;

[0089] Locally recovered abnormal segments are determined based on the changes in local deformation recovery characteristics between adjacent scan locations;

[0090] The abnormal sections of continuous surface undulation, local support, and local recovery are spatially superimposed to obtain candidate tactile abnormality sections. The tactile detection abnormality region is then determined based on the continuous length, abnormal intensity, and location clustering relationship of the candidate tactile abnormality sections.

[0091] Specifically, surface smoothness features are extracted based on the contact pressure data. Surface smoothness features are used to characterize the degree of pressure fluctuation and contact undulation of the seat surface at different scanning positions. When there are bulges, local protrusions, or protruding seams in a certain area, its contact pressure change curve is usually different from that of adjacent areas. By performing position correlation analysis, neighborhood change analysis, and curve fluctuation analysis on the contact pressure data, surface smoothness features that characterize the surface smoothness can be generated.

[0092] Furthermore, soft and hard anomaly features are extracted based on the displacement feedback data. These features characterize the degree of displacement change and support difference of the seat surface at different scanning positions. When a certain area is filled too densely, piled up, has voids, or has local support imbalance, its displacement feedback state will show significant differences under the same preset contact force. By performing distribution analysis and difference analysis on the displacement feedback data, soft and hard anomaly features reflecting local soft and hard unevenness and support anomalies can be generated.

[0093] Furthermore, local deformation recovery features are extracted based on the surface rebound data. These features characterize the recovery time, amplitude, and stability of the seat surface after the contact force is released following pressure. If an area has internal structural anomalies, material fatigue, or foreign matter inclusions, its rebound process may exhibit delayed recovery, insufficient recovery, or abnormal recovery fluctuations. By performing time process analysis and recovery consistency analysis on the surface rebound data, local deformation recovery features can be extracted.

[0094] After the above three features are extracted, the surface flatness features, soft and hard anomaly features, and local deformation recovery features are jointly analyzed to obtain tactile anomaly scores corresponding to different scanning positions, and the tactile detection anomaly areas are determined based on the tactile anomaly scores. The output of this step includes not only feature-level results, but also tactile detection anomaly areas after spatial localization, providing a clear input range for subsequent visual verification.

[0095] S300 performs visual image acquisition on abnormal areas obtained by tactile detection, extracts texture and color abnormality features of the corresponding areas, and generates visual verification results.

[0096] In the above steps, the input content is the tactile detection abnormal area output by S200; the purpose of this step is to perform local fine visual verification on the abnormal areas that have been screened out by tactile detection, so as to reduce the burden of global visual detection and improve the ability to identify surface abnormality types.

[0097] S300 specifically includes the following steps:

[0098] The visual image acquisition center, acquisition angle, and acquisition distance are determined based on the range information of the abnormal tactile detection area.

[0099] The visual acquisition device is controlled to perform directional visual image acquisition on the abnormal tactile detection area to obtain the corresponding local area image;

[0100] Based on the local region image, extract texture anomaly features that characterize the continuous state of surface texture, and color anomaly features that characterize the state of surface color change.

[0101] The texture anomaly features and color anomaly features are mapped to the spatial locations of the tactile detection anomaly areas to generate the visual verification result.

[0102] The visual verification results are generated in the following ways:

[0103] Based on the texture anomaly features, calculate the degree of texture continuity change, texture breakage, and texture misalignment within the tactile detection anomaly area;

[0104] Based on the aforementioned color anomaly characteristics, calculate the degree of color shift, the degree of uneven color distribution, and the degree of color abrupt change within the abnormal tactile detection area;

[0105] The degree of continuous variation in texture, degree of texture breakage, degree of texture misalignment, degree of color shift, degree of uneven color distribution, and degree of color mutation are combined to obtain the visual anomaly score.

[0106] Based on the comparison between the visual anomaly score and the preset visual verification threshold, a visual verification result corresponding to the tactile detection anomaly area is generated.

[0107] Specifically, the visual image acquisition center, acquisition angle, and acquisition distance are first determined based on the range information of the tactile detection abnormal area. The range information of the tactile detection abnormal area can be composed of the boundary of the abnormal segment, continuous length, cluster position, or initial coordinates. By setting the visual image acquisition center, acquisition angle, and acquisition distance based on this information, the visual acquisition device can acquire local area images in a way that adapts to the spatial state of the abnormal area.

[0108] Furthermore, the visual acquisition device is controlled to perform directional visual image acquisition on the tactile detection abnormal area to obtain the corresponding local area image; the visual acquisition device can be an industrial camera, a structured light-assisted camera, or other imaging unit capable of outputting high-resolution local area images; during acquisition, the supplementary light intensity and exposure conditions can be set according to the seat surface material, reflectivity, and seam distribution to improve the stability of texture abnormality feature and color abnormality feature extraction.

[0109] Furthermore, based on the local region image, texture anomaly features representing the continuous state of surface texture and color anomaly features representing the state of surface color change are extracted; texture anomaly features are used to reflect information such as the degree of continuous change of surface texture, the degree of texture breakage and the degree of texture misalignment; color anomaly features are used to reflect information such as the degree of color shift, the degree of uneven color distribution and the degree of color abrupt change.

[0110] Finally, the texture anomaly features and color anomaly features are mapped to the spatial location of the tactile detection anomaly area to generate the visual verification result; the output of this step is the visual verification result, which includes both feature analysis results and spatial correspondence with the tactile detection anomaly area.

[0111] The S400 comprehensively assesses defects such as bulges, depressions, foreign object inclusions, surface wear, and uneven stitching on the seat surface based on tactile detection results and visual verification results, and outputs the corresponding defect categories and defect locations.

[0112] In the above steps, the inputs are tactile detection results and visual verification results. The tactile detection results include at least surface flatness features, softness and hardness anomaly features, local deformation recovery features, and tactile detection abnormal areas. The visual verification results include at least texture anomaly features, color anomaly features, and the image location of the abnormal area. The goal of this step is to distinguish and determine different defect types through multi-feature fusion and rule matching, and output the corresponding defect category and defect location.

[0113] The S400 specifically includes the following steps:

[0114] The surface smoothness features, softness and hardness anomaly features, and local deformation recovery features in the tactile detection results are fused with the texture anomaly features and color anomaly features in the visual verification results to generate defect candidate judgment results.

[0115] Based on the combination relationship of various features in the defect candidate determination results, matching analysis is performed on defects such as bulges, collapses, foreign matter inclusions, surface wear, and uneven stitching.

[0116] When the surface smoothness features are characterized by local bulges and abnormal local deformation recovery features, it is judged as a bulge; when the surface smoothness features are characterized by local depressions and abnormal softness / hardness features, it is judged as a collapse.

[0117] When abnormal texture and color features are localized to be different from the surrounding area, they are identified as foreign matter inclusions. When abnormal texture features are characterized by missing or disordered surface texture, they are identified as surface wear. When abnormal texture features are characterized by abnormal continuity in the suture line area, they are identified as uneven suture defects.

[0118] Specifically, the surface smoothness features, softness and hardness anomaly features, and local deformation recovery features in the tactile detection results are first fused with the texture anomaly features and color anomaly features in the visual verification results to generate defect candidate judgment results. Here, the corresponding fusion is to establish a mapping relationship between tactile response features and visual anomaly features in the same spatial area to form a multi-source joint description result.

[0119] Furthermore, based on the combination relationship of various features in the defect candidate determination results, matching analysis is performed on bulges, collapses, foreign matter inclusions, surface wear, and uneven seams. Specifically, when the surface smoothness feature is characterized by local bulges and abnormal local deformation recovery features, it is determined to be a bulge; when the surface smoothness feature is characterized by local depressions and abnormal softness / hardness features, it is determined to be a collapse; when the texture and color abnormalities are locally different from the surrounding area, it is determined to be a foreign matter inclusion; when the texture abnormalities are characterized by missing or disturbed surface texture, it is determined to be surface wear; when the texture abnormalities are characterized by abnormal continuity in the seam area, it is determined to be an uneven seam defect.

[0120] The output method for the defect location includes:

[0121] The initial position of the tactile detection abnormal area is determined based on the scanning position coordinates corresponding to the movement of the tactile scanning actuator along the preset detection trajectory.

[0122] The image position of the visual verification result is determined based on the imaging area coordinates corresponding to the visual image acquisition device when it acquires visual images of the tactile detection abnormal area.

[0123] By performing coordinate mapping and region alignment between the initial position and the image position, the target position of the corresponding defect on the seat surface is obtained;

[0124] The target location is associated with the defect category and output to form the defect location;

[0125] After the defect category is determined, coordinate mapping and region alignment are performed based on the scan position coordinates and imaging area coordinates to obtain the target position. The target position is then associated with the defect category to form the defect position. The output of this step is the defect category and defect position, which can be directly used for rework location, alarm output, and detection record storage.

[0126] In some embodiments, the seat surface can be divided into a central pressure-bearing area, an edge transition area, and a seam-sensitive area. For the central pressure-bearing area, the scanning step distance can be appropriately increased while maintaining a moderate contact depth to improve the efficiency of large-area detection. For the edge transition area and the seam-sensitive area, the scanning step distance can be reduced and the moving speed can be decreased to improve the recognition accuracy of local undulation abnormalities and seam unevenness defects. If the surface flatness feature is detected as a local bulge in the central pressure-bearing area, and the local deformation recovery feature is abnormal, while the visual verification results show that the texture abnormality feature and color abnormality feature are not obvious, the comprehensive judgment module can determine that the area is a bulge defect. If local depression and abnormal softness / hardness feature are detected in the edge transition area, it can be determined as a collapse defect. If abnormal texture continuity and texture misalignment are detected in the seam-sensitive area, it can be determined as a seam unevenness defect.

[0127] This application also includes an anomaly review and update step, which includes:

[0128] When the tactile detection result is inconsistent with the visual verification result, a second tactile scan is performed on the corresponding tactile detection abnormal area.

[0129] Based on the secondary tactile scan, the contact pressure data, displacement feedback data and surface rebound data of the corresponding positions are re-acquired, and the surface flatness features, soft and hard anomaly features and local deformation recovery features are re-extracted.

[0130] Visual images of the tactile detection abnormal areas are re-acquired, and the texture and color abnormal features of the corresponding areas are re-extracted.

[0131] Based on the re-extracted surface flatness features, softness and hardness anomaly features, local deformation recovery features, texture anomaly features, and color anomaly features, the defects such as bulges, collapses, foreign matter inclusions, surface wear, and uneven seams are updated and judged, and the corresponding defect categories and defect locations are re-output.

[0132] In other embodiments, when the tactile detection result shows a local anomaly, but the visual verification result does not reach the preset visual verification threshold, the system can trigger an anomaly verification update step, perform a second tactile scan of the area with a smaller scanning step, and re-acquire the visual image at a closer acquisition distance. If the re-extracted surface flatness features, softness / hardness anomaly features, and local deformation recovery features still indicate a local anomaly, and the re-extracted texture anomaly features and color anomaly features are still not significant, then the system can be preferentially judged as an internal support type defect based on the tactile detection result. If a significant color change and texture breakage degree anomaly are observed after the second visual verification, then the comprehensive color can be combined to identify a surface anomaly type defect. In this way, the system can improve the reliability of judgment in complex boundary areas and multi-factor interference areas.

[0133] In summary, the seat surface defect detection method and device of the present invention, by constructing a tactile scanning process based on a preset detection trajectory, acquires contact pressure data, displacement feedback data, and surface rebound data, and then extracts surface flatness features, softness and hardness anomaly features, and local deformation recovery features. Based on this, it determines the tactile detection abnormal area, and then combines the texture anomaly features and color anomaly features in the local area image to generate a visual verification result. Finally, it comprehensively judges bulges, collapses, foreign object inclusions, surface wear, and uneven stitching defects and outputs the defect category and defect location. This can effectively improve the automation level, recognition accuracy, and engineering applicability of seat surface defect detection.

[0134] like Figure 2 As shown, a seat surface defect detection device includes:

[0135] The tactile scanning module 10 controls the tactile scanning actuator to move along a preset detection trajectory on the seat surface, perform contact scanning on the seat surface, and acquire contact pressure data, displacement feedback data and surface rebound data at the corresponding positions.

[0136] The tactile feature analysis module 20 extracts surface flatness features, softness / hardness anomaly features, and local deformation recovery features of the seat surface based on the contact pressure data, displacement feedback data, and surface rebound data, and determines the tactile detection abnormal area based on the surface flatness features, softness / hardness anomaly features, and local deformation recovery features.

[0137] The visual verification module 30 acquires visual images of the abnormal areas obtained by tactile detection, extracts the texture and color abnormal features of the corresponding areas, and generates visual verification results.

[0138] The comprehensive judgment module 40, based on the tactile detection results and visual verification results, comprehensively judges the defects on the seat surface, such as bulges, depressions, foreign object inclusions, surface wear, and uneven stitching, and outputs the corresponding defect category and defect location.

[0139] It should be noted that all directional indications in the embodiments of the present invention, such as up, down, left, right, front, back, etc., are only used to explain the relative positional relationship and movement of the components in a specific posture, as shown in the attached figure. If the specific posture changes, the directional indication will also change accordingly.

[0140] Furthermore, the use of terms such as "first" and "second" in this invention is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Meanwhile, the word "and / or" throughout the text means including three solutions; for example, "A and / or B" includes solution A, solution B, or a solution that simultaneously satisfies A and B. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0141] All of the above components are general standard parts or components known to those skilled in the art. Their structure and principles can be learned by those skilled in the art through technical manuals or conventional experimental methods.

[0142] For example, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the electronic device.

[0143] The electronic device may be a desktop computer, laptop, handheld computer, or smart tablet, etc. The electronic device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that the above components are merely examples of electronic devices and do not constitute a limitation on the electronic device. It may include more or fewer components than described above, or combine certain components, or different components. For example, the electronic device may also include input / output devices, network access devices, buses, etc.

[0144] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the electronic device, connecting all parts of the electronic device via various interfaces and lines.

[0145] The memory can be used to store the computer programs and / or modules. The processor implements various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory and by calling data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0146] Wherein, if the modules / units integrated in the electronic device are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0147] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0148] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for detecting surface defects in a seat, characterized in that, Includes the following steps: The seat surface is scanned for contact, and contact pressure data, displacement feedback data and surface rebound data at the corresponding locations are obtained. Based on the contact pressure data, displacement feedback data, and surface rebound data, the surface flatness features, softness / hardness anomaly features, and local deformation recovery features of the seat surface are extracted, and the tactile detection abnormal area is determined based on the surface flatness features, softness / hardness anomaly features, and local deformation recovery features. Visual images are acquired from the abnormal areas obtained by tactile detection, and the texture and color abnormal features of the corresponding areas are extracted to generate visual verification results. Based on the tactile detection results and visual verification results, the defects of the seat surface, such as bulges, depressions, foreign object inclusions, surface wear, and uneven stitching, are comprehensively judged, and the corresponding defect categories and defect locations are output.

2. The method for detecting surface defects of a seat according to claim 1, characterized in that, The step of performing contact scanning on the seat surface and acquiring contact pressure data, displacement feedback data, and surface rebound data at the corresponding locations includes the following steps: Obtain the contour boundary information, surface undulation information, and partition position information of the seat surface, and generate the preset detection trajectory based on the contour boundary information, surface undulation information, and partition position information; The corresponding scanning step distance, contact depth, and moving speed are set according to the curvature and boundary changes of different areas of the seat surface; The tactile scanning actuator is controlled to move point by point along the preset detection trajectory, and a preset contact force is applied to the seat surface at each scanning position; During the movement of the tactile scanning actuator along the preset detection trajectory, contact pressure data, displacement feedback data, and surface rebound data corresponding to each scanning position are collected simultaneously, and the correspondence between the contact pressure data, displacement feedback data, and surface rebound data and the scanning position is established.

3. The method for detecting surface defects of a seat according to claim 1, characterized in that, The step of extracting surface flatness features, softness / hardness anomaly features, and local deformation recovery features of the seat surface based on the contact pressure data, displacement feedback data, and surface rebound data, and determining tactile detection abnormal areas based on the surface flatness features, softness / hardness anomaly features, and local deformation recovery features, includes the following steps: The pressure fluctuation and contact undulation of the seat surface at different scanning positions are extracted based on the contact pressure data to generate the surface flatness feature. The displacement change and support difference of the seat surface at different scanning positions are extracted based on the displacement feedback data to generate the soft and hard anomaly features. Based on the surface rebound data, the recovery time, recovery amplitude, and recovery stability of the seat surface after being compressed are extracted to generate the local deformation recovery characteristics. By jointly analyzing the surface smoothness features, softness and hardness anomaly features, and local deformation recovery features, tactile anomaly scores corresponding to different scanning positions are obtained, and the tactile detection anomaly areas are determined based on the tactile anomaly scores.

4. The method for detecting surface defects of a seat according to claim 3, characterized in that, The methods for determining abnormal tactile detection areas based on the surface smoothness characteristics, softness / hardness anomaly characteristics, and local deformation recovery characteristics include: The abnormal sections of continuous surface undulations are determined based on the changes in surface smoothness characteristics between adjacent scanning positions. Local support anomaly segments are determined based on the changes in soft and hard anomaly characteristics between adjacent scan locations; Locally recovered abnormal segments are determined based on the changes in local deformation recovery characteristics between adjacent scan locations; The abnormal sections of continuous surface undulation, local support, and local recovery are spatially superimposed to obtain candidate tactile abnormality sections. The tactile detection abnormality region is then determined based on the continuous length, abnormal intensity, and location clustering relationship of the candidate tactile abnormality sections.

5. The method for detecting surface defects of a seat according to claim 1, characterized in that, The process of acquiring visual images of abnormal areas detected by tactile sensing, extracting texture and color anomaly features of the corresponding areas, and generating visual verification results includes the following steps: The visual image acquisition center, acquisition angle, and acquisition distance are determined based on the range information of the abnormal tactile detection area. The visual acquisition device is controlled to perform directional visual image acquisition on the abnormal tactile detection area to obtain the corresponding local area image; Based on the local region image, extract texture anomaly features that characterize the continuous state of surface texture, and color anomaly features that characterize the state of surface color change. The texture anomaly features and color anomaly features are mapped to the spatial locations of the tactile detection anomaly areas to generate the visual verification result.

6. The method for detecting surface defects of a seat according to claim 5, characterized in that, The visual verification results are generated in the following ways: Based on the texture anomaly features, calculate the degree of texture continuity change, texture breakage, and texture misalignment within the tactile detection anomaly area; Based on the aforementioned color anomaly characteristics, calculate the degree of color shift, the degree of uneven color distribution, and the degree of color abrupt change within the abnormal tactile detection area; The degree of continuous variation in texture, degree of texture breakage, degree of texture misalignment, degree of color shift, degree of uneven color distribution, and degree of color mutation are combined to obtain the visual anomaly score. Based on the comparison between the visual anomaly score and the preset visual verification threshold, a visual verification result corresponding to the tactile detection anomaly area is generated.

7. The method for detecting surface defects of a seat according to claim 1, characterized in that, Based on the tactile detection results and visual verification results, the defects on the seat surface, such as bulges, depressions, foreign object inclusions, surface wear, and uneven stitching, are comprehensively judged, and the corresponding defect categories and locations are output. This includes the following steps: The surface smoothness features, softness and hardness anomaly features, and local deformation recovery features in the tactile detection results are fused with the texture anomaly features and color anomaly features in the visual verification results to generate defect candidate judgment results. Based on the combination relationship of various features in the defect candidate determination results, matching analysis is performed on defects such as bulges, collapses, foreign matter inclusions, surface wear, and uneven stitching. When the surface smoothness features are characterized by local bulges and abnormal local deformation recovery features, it is judged as a bulge; when the surface smoothness features are characterized by local depressions and abnormal softness / hardness features, it is judged as a collapse. When abnormal texture and color features are localized to be different from the surrounding area, they are identified as foreign matter inclusions. When abnormal texture features are characterized by missing or disordered surface texture, they are identified as surface wear. When abnormal texture features are characterized by abnormal continuity in the suture line area, they are identified as uneven suture defects.

8. The method for detecting surface defects of a seat according to claim 7, characterized in that, The output method for the defect location includes: The initial position of the tactile detection abnormal area is determined based on the scanning position coordinates corresponding to the movement of the tactile scanning actuator along the preset detection trajectory. The image position of the visual verification result is determined based on the imaging area coordinates corresponding to the visual image acquisition device when it acquires visual images of the tactile detection abnormal area. By performing coordinate mapping and region alignment between the initial position and the image position, the target position of the corresponding defect on the seat surface is obtained; The target location is associated with the defect category and output to form the defect location.

9. The method for detecting surface defects of a seat according to claim 8, characterized in that, It also includes an anomaly review and update step, which includes: When the tactile detection result is inconsistent with the visual verification result, a second tactile scan is performed on the corresponding tactile detection abnormal area. Based on the secondary tactile scan, the contact pressure data, displacement feedback data and surface rebound data of the corresponding positions are re-acquired, and the surface flatness features, soft and hard anomaly features and local deformation recovery features are re-extracted. Visual images of the tactile detection abnormal areas are re-acquired, and the texture and color abnormal features of the corresponding areas are re-extracted. Based on the re-extracted surface flatness features, softness and hardness anomaly features, local deformation recovery features, texture anomaly features, and color anomaly features, the defects such as bulges, collapses, foreign matter inclusions, surface wear, and uneven seams are updated and judged, and the corresponding defect categories and defect locations are re-output.

10. A seat surface defect detection device, comprising: The tactile scanning module controls the tactile scanning actuator to move along a preset detection trajectory on the seat surface, perform contact scanning on the seat surface, and acquire contact pressure data, displacement feedback data, and surface rebound data at the corresponding positions. The tactile feature analysis module extracts surface flatness features, softness / hardness anomaly features, and local deformation recovery features of the seat surface based on the contact pressure data, displacement feedback data, and surface rebound data, and determines the tactile detection abnormal area based on the surface flatness features, softness / hardness anomaly features, and local deformation recovery features. The visual verification module acquires visual images of abnormal areas obtained by tactile detection, extracts texture and color abnormality features of the corresponding areas, and generates visual verification results. The comprehensive judgment module, based on the tactile detection results and visual verification results, comprehensively judges the defects on the seat surface such as bulges, depressions, foreign object inclusions, surface wear, and uneven stitching, and outputs the corresponding defect category and defect location.

Images