An underwear deformation degree detection method, device, equipment and storage medium

Abstract

The present application relates to underwear shape change detection technology field, especially in underwear shape change degree detection method, device, equipment and storage medium;Based on the preset B spline surface interpolation algorithm, the global mark point data set is fitted to get the underwear surface model;According to the preset reference curvature parameter, the preset deformation index calculation formula and the preset deformation analysis method, the global high definition reference image and the underwear surface model are deformed to get the comprehensive deformation index;According to the preset quantification rating method, the laser profile scanning surface data, the segmented fatigue detection data, the motion time series data, the stress time series data and the comprehensive deformation index are analyzed to get the bending deformation grade;Through the fusion of multidimensional dynamic stress data and the setting of defect threshold grading, all kinds of deformation defects are covered, and the underwear deformation is refined, standardized and quantified.

Description

Technical Field

[0001] This invention relates to the field of underwear deformation detection technology, and in particular to a method, apparatus, equipment and storage medium for detecting the degree of underwear deformation. Background Technology

[0002] Currently, most deformation detection technologies for underwear fabrics rely on manual visual assessment or single three-dimensional surface comparison, resulting in relatively crude detection methods and limited evaluation dimensions. Existing technologies struggle to comprehensively simulate various defects in underwear caused by stretching, compression, and repeated fatigue stress during wear, including static deformation, dynamic hysteresis deformation, residual plastic deformation, and wrinkle deformation. They also cannot distinguish between instantaneous elastic deformation and permanent stress damage. Furthermore, existing surface reconstruction methods are prone to contour distortion and loss of detail, making it difficult to effectively replicate the true surface morphology after stress loading. Moreover, the lack of unified evaluation standards for various deformation parameters and the inability to couple and analyze deformation data from different working conditions makes it difficult to objectively quantify and assess the bending deformation, fatigue rebound attenuation, and dynamic deformation hysteresis quality of underwear. This leads to low repeatability and standardization, failing to meet the industrial application needs for refined quality testing and grading evaluation of high-end underwear. Summary of the Invention

[0003] In order to overcome the shortcomings of the prior art, the purpose of this invention is to provide a method, apparatus, device and storage medium for detecting the degree of deformation of underwear.

[0004] A method for detecting the degree of deformation of underwear includes: acquiring a full-area high-definition reference image and a full-area marker point dataset of the underwear; performing surface fitting on the full-area marker point dataset based on a preset B-spline surface interpolation algorithm to obtain an underwear surface model; performing deformation analysis on the full-area high-definition reference image and the underwear surface model according to preset reference curvature parameters, preset deformation index calculation formula, and preset deformation analysis method to obtain a comprehensive deformation index; acquiring laser contour scanning surface data, segmented fatigue detection data, motion time series data, and stress time series data of the scanning area; and performing rating analysis on the laser contour scanning surface data, segmented fatigue detection data, motion time series data, stress time series data, and comprehensive deformation index according to a preset quantitative rating method to obtain the bending deformation level.

[0005] Furthermore, the step of performing surface fitting on the global marked point dataset based on a preset B-spline surface interpolation algorithm to obtain an underwear surface model includes: performing search analysis on the global marked point dataset according to a preset neighborhood search algorithm and a preset radius range to obtain a sparse marked point dataset and a continuous marked point dataset; and performing surface fitting on the sparse marked point dataset and the continuous marked point dataset based on the B-spline surface interpolation algorithm to obtain an underwear surface model.

[0006] Furthermore, the step of performing deformation analysis on the full-domain high-definition reference image and the underwear surface model according to preset reference curvature parameters, preset deformation index calculation formula, and preset deformation analysis method to obtain a comprehensive deformation index includes: extracting features from the full-domain high-definition reference image according to a preset edge detection algorithm to obtain edge features and surface contours; dividing the full-domain high-definition reference image into regions according to edge features and surface contours to obtain multiple detection zones; performing deformation analysis on the multiple detection zones and the underwear surface model according to the reference curvature parameters and deformation analysis method to obtain static deformation rate, dynamic deformation rate, residual deformation rate, and wrinkle deformation rate; and calculating the static deformation rate, dynamic deformation rate, residual deformation rate, and wrinkle deformation rate according to the deformation index calculation formula to obtain a comprehensive deformation index.

[0007] Furthermore, the step of rating and analyzing the laser profile scanning surface data, segmented fatigue detection data, motion time series data, stress time series data, and comprehensive deformation index according to a preset quantitative rating method to obtain the bending deformation level includes: analyzing the laser profile scanning surface data and segmented fatigue detection data according to a preset standard fitted elastic rebound attenuation curve and a preset rebound performance calculation formula to obtain the rebound performance index; and rating and analyzing the motion time series data, stress time series data, comprehensive deformation index, and rebound performance index according to the quantitative rating method to obtain the bending deformation level.

[0008] Furthermore, the process of analyzing the laser profile scanning surface data and segmented fatigue detection data based on a preset standard fitted elastic rebound attenuation curve and a preset elastic rebound performance calculation formula to obtain the elastic rebound performance index includes: Point cloud registration is performed on the underwear surface model based on preset reference curvature parameters, preset reference surfaces, and preset minimum distance iterative matching algorithms to obtain the surface curvature fitting deformation characteristics; the surface curvature fitting deformation characteristics and laser contour scanning surface data are fitted and analyzed based on preset curvature fitting algorithms to obtain the degree of bending deformation; the degree of bending deformation and segmented fatigue detection data are analyzed based on standard fitted elastic rebound attenuation curves and elasticity calculation formulas to obtain the elasticity performance index.

[0009] Furthermore, the step of analyzing the bending deformation degree and segmented fatigue test data according to the standard fitted elastic rebound attenuation curve and the rebound performance calculation formula to obtain the rebound performance index includes: performing attenuation analysis on the bending deformation degree and segmented fatigue test data according to the standard fitted elastic rebound attenuation curve to obtain the fatigue attenuation coefficient; and calculating the fatigue attenuation coefficient according to the rebound performance calculation formula to obtain the rebound performance index.

[0010] Furthermore, the step of performing rating analysis on motion time series data, stress time series data, comprehensive deformation index, and rebound performance index according to the quantitative rating method to obtain the bending deformation level includes: performing joint feature analysis on motion time series data and stress time series data according to a preset multi-source time series cross-correlation feature matching analysis method to obtain the deformation response time difference and deformation hysteresis; and performing rating analysis on the comprehensive deformation index, deformation response time difference, deformation hysteresis, and rebound performance index according to the quantitative rating method to obtain the deformation degree level.

[0011] Furthermore, a device for detecting the degree of deformation of underwear includes: a data acquisition module for acquiring a full-area high-definition reference image and a full-area marker point dataset of the underwear; a surface fitting module for performing surface fitting on the full-area marker point dataset based on a preset B-spline surface interpolation algorithm to obtain an underwear surface model; a deformation analysis module for performing deformation analysis on the full-area high-definition reference image and the underwear surface model according to preset reference curvature parameters, preset deformation index calculation formulas, and preset deformation analysis methods to obtain a comprehensive deformation index; a data acquisition module for acquiring laser contour scanning surface data, segmented fatigue detection data, motion time series data, and stress time series data of the scanning area; and a rating analysis module for performing rating analysis on the laser contour scanning surface data, segmented fatigue detection data, motion time series data, stress time series data, and comprehensive deformation index according to a preset quantitative rating method to obtain a bending deformation level.

[0012] Furthermore, an underwear deformation degree detection device includes: a memory and at least one processor, wherein the memory stores instructions; the at least one processor invokes the instructions in the memory to cause the underwear deformation degree detection device to perform the various steps of the underwear deformation degree detection method as described above.

[0013] Furthermore, a computer-readable storage medium stores instructions that, when executed by a processor, implement the steps of the underwear deformation detection method described above.

[0014] In this invention, a comprehensive quantitative inspection of underwear subjected to stress deformation is conducted. Images and marker data with realistic tensile and compressive deformation characteristics are collected under standardized working conditions to ensure the uniformity and authenticity of the sample data. A constrained B-spline surface interpolation algorithm is used to reconstruct a three-dimensional deformation surface, replicating the contour deformation and local wrinkle curvature of the underwear under stress, effectively solving the problems of edge distortion and lack of detail in traditional surface reconstruction. Based on the factory-set stress-free reference curvature, and integrating the principles of two-dimensional pixel difference and three-dimensional curvature differentiation, four types of stress deformation rates—static, dynamic, residual, and wrinkle—are quantified in different regions. A comprehensive deformation index, calculated through weighted normalization, comprehensively characterizes the degree of stress damage to the underwear. Simultaneously, multi-dimensional dynamic data of fatigue cyclic stress and human body alternating motion stress are collected, and a defect threshold correction mechanism is used to complete multi-level quantitative rating. This solution takes into account instantaneous stress deformation, permanent plastic damage, and long-term fatigue decay, comprehensively covering various deformation defects in underwear, effectively improving the precision, objectivity, and standardization of stress deformation detection, and achieving quantitative rating of the bending deformation quality of underwear. Attached Figure Description

[0015] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 A flowchart of a method for detecting the degree of deformation of underwear provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of a device for detecting the degree of deformation of underwear provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an underwear deformation detection device provided in an embodiment of the present invention. Detailed Implementation

[0016] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0017] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1One embodiment of the method for detecting the degree of deformation of underwear in this invention includes: 101. Collect full-domain high-resolution baseline images and full-domain marker point datasets of underwear; In this embodiment, raw data of the underwear under test, which has undergone deformation after being subjected to tensile and compressive stresses, is collected. The overall acquisition logic is to obtain images and point data of deformation under actual wearing stress as the benchmark for the sample under test for deformation analysis. The acquisition process must simultaneously meet two basic constraints: First, the underwear must undergo standard stress cyclic loading in advance to maintain a constant stress state that simulates human wear and ensures that the surface deformation shape is fixed without rebound fluctuations. Second, a D65 standard light source, a fixed shooting distance, and a three-dimensional scanning sampling density are uniformly used to simultaneously complete the two-dimensional full-domain high-definition image capture and three-dimensional spatial marker point scanning and registration, ensuring that the image pixel scale and the global coordinates of the marker points have a unified comparison standard. Simultaneously, a full-domain high-definition benchmark image with stress deformation characteristics and a full-domain marker point dataset under the same coordinate system are output, providing the original input basis for the deformation under test for subsequent three-dimensional surface reconstruction, structural zoning, and stress deformation analysis. 102. Based on the preset B-spline surface interpolation algorithm, surface fitting is performed on the global marked point dataset to obtain the underwear surface model; In this embodiment, after the complete global marker point dataset with stress deformation characteristics is collected, the system first uses a fixed radius neighborhood search algorithm to split the data into two subsets: sparse marker points at the boundary and continuous dense marker points inside. Then, it calls a preset bicubic B-spline surface interpolation algorithm to perform a constraint-based fitting operation. The overall surface reconstruction logic is to simultaneously lock the true contour boundary after stress deformation and restore the details of internal stress bending and wrinkling. The fitting process must simultaneously satisfy two basic constraints: first, sparse marker points are forced boundary value points, and the fitted surface must pass through the entire boundary. The system employs two main methods: first, it replicates the deformation of the outer contour of the underwear under stress, avoiding edge shrinkage and contour distortion; second, it uses continuous and dense marker points as internal interpolation points to fill the middle area of ​​the surface, restoring the indentation and stretching curvature of the cups, side wings, and elastic bands caused by stress; and third, it uses B-spline control vertices and node vectors to continuously optimize the curvature continuity of the surface, generating a complete, unbroken, and realistic three-dimensional model of the underwear surface to be tested, providing a three-dimensional calculation basis for subsequent registration and comparison with the factory stress-free standard model and curvature deformation analysis. 103. Perform deformation analysis on the full-domain high-definition reference image and the underwear surface model according to the preset reference curvature parameters, the preset deformation index calculation formula and the preset deformation analysis method to obtain the comprehensive deformation index. In this embodiment, edge features and closed surface contours are extracted based on high-definition reference images of stress deformation across the entire domain. The contour boundaries are used to automatically partition the stress structure of the underwear, generating multiple sets of standardized independent detection partitions that do not overlap and have complete boundaries. The system retrieves the factory-specific stress-free reference curvature parameters and the full-domain deformation analysis method for each partition, and synchronously couples and compares the two-dimensional image under the current stress state, the three-dimensional surface model of stress deformation, and the factory-specific zero-stress standard model. The overall deformation analysis logic is to quantify the four types of differentiated deformation defects induced by stress in different dimensions. The analysis process must simultaneously meet two basic constraints: first, each detection partition must match the factory-specific reference curvature as the arc. First, by comparing the scale, the two-dimensional pixel deformation and three-dimensional surface curvature deformation data are calculated independently in partitions, without cross-regional mixed calculations. Second, by integrating the dual principles of two-dimensional pixel coordinate difference matching and three-dimensional surface differential curvature analysis, the static deformation rate, dynamic peak deformation rate, permanent residual deformation rate after pressure relief, and surface wrinkle and depression deformation rate are solved respectively by referring to the stress-free benchmark. After normalizing the range of all deformation rates, they are substituted into the deformation index formula that distinguishes the defect weights for weighted fusion, and the dimensionless comprehensive deformation index in the range of 0 to 1 is output. This uniformly represents the severity of the full-domain deformation defects of underwear after being subjected to stress, and provides core evaluation parameters for stress deformation for multi-dimensional coupled quantitative rating. 104. Acquire laser contour scanning surface data, segmented fatigue detection data, motion timing data, and force timing data of the scanning area; In this embodiment, after the static stress comprehensive deformation index is solved, a multi-condition synchronous data acquisition process is initiated to fully cover two types of dynamic evaluation scenarios: long-term cyclic fatigue stress and dynamic alternating motion stress of the human body. The overall acquisition logic is to supplement the long-term fatigue loss and dynamic instantaneous alternating deformation characteristic data in addition to the single static stress deformation. The acquisition process must simultaneously meet two basic constraints: First, segmented fatigue detection is carried out at fixed cyclic stress intervals of 10,000, 20,000, 30,000, 40,000, and 50,000 cycles to collect deformation steady-state samples and remove abnormal fluctuation data caused by stress mutations during the start and stop phases of fatigue testing. Second, motion timing and stress timing are synchronously acquired using a unified millisecond-level timestamp to simulate the alternating stress applied by human body activity cycles. After filtering to remove equipment jitter noise, only the effective motion stress cycle data of rotation, arm raising, and flexion and extension are retained. The high-precision laser contour surface under continuous stress, multi-level fatigue segmented stress detection, and synchronous motion stress timing data are output synchronously to support the solution of the rebound performance index and the quantification calculation of dynamic deformation response time difference and hysteresis, respectively. 105. Based on the preset quantitative rating method, the laser profile scanning surface data, segmented fatigue detection data, motion time series data, stress time series data and comprehensive deformation index are rated and analyzed to obtain the bending deformation level. In this embodiment, after collecting all evaluation parameters including static stress comprehensive deformation index, stress laser profile, fatigue segmented stress data, and dynamic alternating stress time sequence, the system calls a preset quantitative rating method to conduct multi-dimensional fusion and grading judgment. The overall rating logic is to take into account the single static stress fit deviation, long-term cyclic fatigue stress rebound attenuation, and human body alternating motion stress deformation lag deformation damage under all working conditions to achieve objective and standardized grading. The judgment process must simultaneously meet two basic constraints: First, the stress deformation response time difference and lag amount are solved by multi-source cross-correlation feature matching of motion and stress time sequence, and all heterogeneous parameters are uniformly normalized and mapped to the 0-1 range to eliminate the problem of imbalance of different indicators; Second, a single stress defect hard threshold locking correction mechanism is embedded. If the rebound index is too low, the dynamic stress lag exceeds the standard, or the comprehensive deformation index exceeds the limit, the comprehensive score is downgraded and deducted. The comprehensive score after defect correction is matched with the preset five-level deformation threshold range one by one to output the standardized bending deformation level, realizing the integrated objective quantitative rating of underwear single static stress, long-term cyclic fatigue stress, and dynamic alternating motion stress deformation. In this embodiment, a comprehensive quantitative inspection is conducted on underwear subjected to stress deformation. Images and marker data with realistic tensile and compressive deformation characteristics are collected under standardized working conditions to ensure the uniformity and authenticity of the sample data. A constrained B-spline surface interpolation algorithm is used to reconstruct a three-dimensional deformation surface, replicating the contour deformation and local wrinkle curvature of the underwear under stress, effectively solving the problems of edge distortion and lack of detail in traditional surface reconstruction. Based on the factory-set stress-free reference curvature, and integrating the principles of two-dimensional pixel difference and three-dimensional curvature differentiation, four types of stress deformation rates—static, dynamic, residual, and wrinkle—are quantified in different regions. A comprehensive deformation index is obtained through weighted normalization calculation to comprehensively characterize the degree of stress damage to the underwear. Simultaneously, multi-dimensional dynamic data of fatigue cyclic stress and human body alternating motion stress are collected, and a defect threshold correction mechanism is used to complete multi-level quantitative rating. This solution takes into account instantaneous stress deformation, permanent plastic damage, and long-term fatigue decay, comprehensively covering various deformation defects in underwear, effectively improving the precision, objectivity, and standardization of stress deformation detection, and achieving quantitative rating of the bending deformation quality of underwear.

[0018] A second embodiment of a method for detecting the degree of deformation of underwear according to the present invention includes: 201. Perform search and analysis on the global marked point dataset according to the preset neighborhood search algorithm and preset radius range to obtain sparse marked point dataset and continuous marked point dataset; In this embodiment, a preset neighborhood search algorithm and a fixed spatial search radius together constitute the underlying constraints for point classification, used to perform spatial neighborhood statistics and classification filtering on all three-dimensional marker points of the underwear. The overall point division logic is to distinguish between sparse points at the surface boundary and dense continuous points inside the surface. The classification process must simultaneously satisfy two basic constraints: First, all marker points in the entire domain are accompanied by unified three-dimensional spatial coordinates, and the coordinates are uniformly transformed to the global coordinate system of the underwear reference surface to eliminate the interference of coordinate reference offset; Second, a preset fixed spatial radius is used as the neighborhood statistics threshold. When traversing each marker point, only the number of adjacent marker points within the spherical range of this radius is counted. By comparing the number of neighboring points point by point, if the number of points in the neighborhood is lower than the preset neighborhood statistics threshold, it is classified into the boundary sparse marker point dataset; if the number of points in the neighborhood is greater than or equal to the neighborhood statistics threshold, it is classified into the internal continuous dense marker point dataset. Two sets of marker points with independent attributes are obtained, providing a basis for the input of partitioned value points for subsequent constraint B-spline surface interpolation fitting. 202. Based on the B-spline surface interpolation algorithm, surface fitting is performed on sparse marker point datasets and continuous marker point datasets to obtain the underwear surface model; In this embodiment, after classifying and splitting the sparse and continuous marker point datasets, the system calls a preset bicubic B-spline surface interpolation algorithm to perform a constraint-based surface fitting operation. The overall surface reconstruction logic is to simultaneously use sparse boundary points to lock the contour shape and internal continuous points to restore the surface details. The fitting process simultaneously satisfies two basic constraints: First, the sparse marker point dataset is set as interpolation boundary forced constraint value points, and the fitted surface must pass through all boundary sparse points to avoid surface edge shrinkage and contour distortion. Second, the continuous marker point dataset is used as internal interpolation value points to fill the middle area of ​​the surface and restore the smooth curvature of areas such as cups and wings. The global contour boundary is locked by sparse points, and the internal continuous surface is fitted by continuous dense points. The B-spline surface control vertex and node vectors are jointly solved, and the surface smoothness and curvature continuity are iteratively optimized. Finally, a complete, unbroken, and contour-fitting global continuous 3D surface model of underwear is generated, providing a complete 3D base model for subsequent point cloud registration and curvature deformation analysis. In this embodiment, a fixed-radius neighborhood search algorithm is used to intelligently classify the 3D marker points across the entire domain. A unified coordinate system and a neighborhood number threshold are used to distinguish between sparse boundary marker points and dense, continuous internal marker points, achieving structured and differentiated splitting of the marker point dataset. Based on this, a B-spline surface interpolation algorithm is used for constraint fitting. Sparse marker points are used to forcibly lock the surface boundary contour, avoiding edge collapse and contour distortion. Simultaneously, dense, continuous marker points are used to restore the smooth curvature and detailed features of the internal surface of the underwear, ensuring the overall curvature continuity and the true integrity of the shape. Through a differentiated point-position adaptation fitting strategy, the model boundary fidelity and internal surface smoothness are balanced, restoring key structural forms such as the underwear cup, side wings, and elastic bands. This effectively improves the accuracy and structural integrity of the 3D surface model reconstruction, providing a highly reliable 3D model base for subsequent surface registration, curvature analysis, and deformation detection, thus enhancing the overall accuracy and stability of underwear deformation detection.

[0019] A third embodiment of a method for detecting the degree of deformation of underwear according to the present invention includes: 301. Extract features from the full-domain high-definition reference image according to the preset edge detection algorithm to obtain edge features and surface contours; In this embodiment, the preset edge detection algorithm incorporates gradient threshold filtering and morphological denoising as dual underlying constraints to perform pixel-level structural feature analysis on the full-domain high-definition reference image of the underwear. The overall extraction logic involves segmenting the underwear body from the blank background, distinguishing various structural boundaries, and extracting complete structural contour information. The extraction process must simultaneously satisfy two basic constraints: First, the full-domain high-definition reference image undergoes pre-processing illumination uniformity correction to eliminate pixel gradient interference caused by uneven brightness of the D65 light source; second, the edge detection algorithm sets high and low gradient thresholds to retain true weak edges of the structure and filter out subtle fabric texture noise and image graininess. Through row-by-row and column-by-column pixel gradient traversal and connected component filtering, isolated false edges are eliminated, and the edge features and continuous closed surface contours corresponding to the outer contour of the underwear, seams, elastic band boundaries, and cup boundaries are fully output, providing boundary input basis for subsequent image structured partitioning. 302. Divide the full-domain high-definition reference image into regions based on edge features and surface contours to obtain multiple detection zones; In this embodiment, the system performs structured automatic zoning of underwear based on the closed contour boundary. The overall zoning logic is to split the underwear into independent detection units according to the physiological stress structure of the underwear. The zoning process simultaneously meets two basic constraints: First, the closed curved contour is used as the dividing boundary to distinguish the outer boundary of the product from the internal structural dividing line, and the boundary does not overlap or have any blank areas. Second, the zoning rules follow the stress characteristics of underwear, dividing the bra into left cup, right cup, underbust, side wings, and shoulder straps, and dividing panties and shapewear into front piece, back piece, and side waist elastic area. Through the filling of the closed contour area and the grouping and classification of the boundary, multiple standardized detection zones with clear boundaries are automatically obtained, realizing one-to-one correspondence between the curved surface and image data zoning, and defining independent calculation units for the refined deformation analysis of the region. 303. Based on the reference curvature parameters and deformation analysis methods, perform deformation analysis on multiple detection zones and underwear surface models to obtain static deformation rate, dynamic deformation rate, residual deformation rate and wrinkle deformation rate; In this embodiment, deformation analysis integrates the two-dimensional pixel geometric registration deformation principle with the three-dimensional differential surface curvature analysis principle: At the two-dimensional level, based on the image pixel coordinate difference comparison principle, using the stress-free reference image partition contour as a reference, the planar deformation of the fabric plane due to stretching, shrinkage, and stacking wrinkles is calculated by converting the pixel offset distance; at the three-dimensional level, relying on the B-spline surface differential curvature solution principle, the first and second order differential curvatures are obtained from the fitted continuous underwear surface, and the difference is calculated with the partition-specific reference curvature to quantify the three-dimensional curvature deformation caused by surface bending, concavity, and convexity; a unified stress-free factory standard state is used as the zero-deformation reference threshold to distinguish between instantaneous reversible elastic deformation and irreversible permanent plastic deformation, simultaneously integrating two types of information—planar image scale change and three-dimensional surface curvature deviation—to achieve accurate quantification of multi-dimensional deformation; and obtaining structured independent detection scores. After the area is divided, the system retrieves the preset benchmark curvature parameters and the global deformation analysis method to synchronously couple and analyze the two-dimensional images and three-dimensional underwear surface models of each area. The overall deformation analysis logic is to quantify the deformation amplitude of underwear under different working conditions in different dimensions. The analysis process simultaneously meets two basic constraints: First, each detection area is matched with exclusive factory standard benchmark curvature parameters as a unified scale for comparison of arc and size; Second, the pixel deformation of two-dimensional images and the curvature deformation data of three-dimensional surfaces are calculated independently by area, without cross-area mixed calculation. The system solves four types of deformation indices, namely static deformation, dynamic peak deformation, permanent residual deformation after pressure relief, and surface wrinkle and concave deformation, respectively, against the stress-free benchmark state. The system outputs four basic deformation parameters, namely static deformation rate, dynamic deformation rate, residual deformation rate, and wrinkle deformation rate, which provide four types of original deformation input indices for the weighted fusion calculation of the comprehensive deformation index. 304. Calculate the static deformation rate, dynamic deformation rate, residual deformation rate, and fold deformation rate according to the deformation index calculation formula to obtain the comprehensive deformation index; In this embodiment, after all four types of deformation rate parameters for different regions are solved, the system calls the preset deformation index calculation formula to perform multi-index weighted fusion and normalization calculation. The overall fusion logic is to coordinate the four types of deformation defects—static fit, dynamic movement, prolonged wear loosening, and fabric wrinkles—to form a unified quantitative evaluation index. The calculation process simultaneously satisfies two basic constraints: First, the four types of deformation rates are assigned independent fixed weights according to the degree of negative impact on the quality of underwear wearing, with residual deformation and wrinkle deformation assigned higher defect weights; second, all individual deformation rates undergo range normalization to eliminate the evaluation imbalance caused by differences in the numerical ranges of different indicators. The normalized four deformation rates are substituted into the fusion formula for weighted superposition correction, and a unified dimensionless global comprehensive deformation index is output. The higher the value, the more serious the overall deformation defect of the underwear. This index serves as an integrated evaluation index for static, dynamic, wrinkle, and residual deformation, supporting subsequent multi-parameter coupled quantitative rating. The formula for calculating the deformation index is as follows: , In the formula, The comprehensive deformation index is a dimensionless fusion evaluation index with a value range of [0,1]. The closer the value is to 1, the more serious the overall deformation defect of the underwear testing area is, and the worse the comprehensive performance of the four types of problems: static fit deviation, dynamic movement deformation, permanent loosening after long wear, and fabric wrinkles. The closer the value is to 0, the more uniform the deformation of the area, the fewer the deformation defects, and the better the fit performance. A fixed constant is preset for the static deformation weight coefficient, representing the proportion of the static deformation rate's contribution to the comprehensive deformation index; this satisfies the weight constraints. It is the smallest of the four weights, and its value range is... The sum of the four weights Used to mitigate the negative impact of slight static deformation on the overall rating; The normalized static deformation rate is the dimensionless value of the original static deformation rate after range normalization mapping, in the range [0,1]; it represents the average extension and bending deformation amplitude of the partitioned surface relative to the reference stress-free surface when the underwear is in a static and motionless state. A fixed constant is preset for the dynamic deformation weight coefficient, which represents the proportion of the evaluation contribution of the dynamic deformation rate. The weight is only higher than that of the static deformation weight. value range Used to moderately amplify the impact of the instantaneous deformation of underwear on the overall deformation during human movement; The normalized dynamic deformation rate is the dimensionless value of the normalized range of the original dynamic deformation rate, in the interval [0,1]. The normalized result of the maximum deformation amplitude of the partitioned surface under the peak state of simulated human body rotation, arm raising, flexion and extension movements. The residual deformation weight coefficient is a preset fixed constant, with the largest value among the four weights, and its range is [not specified]. Permanent residual deformation is an irreversible wear defect, and is given the highest weight to highlight the negative impact of prolonged wear and permanent deformation on the quality of underwear. The normalized residual deformation rate is obtained by normalizing the original residual deformation rate, with a range of [0,1]. It represents the proportion of permanent irreversible deformation remaining in the underwear after all external forces are removed and it has fully rebounded and stabilized, relative to the reference size / radian. A fixed constant is preset for the fold deformation weighting coefficient, and the weight value is lower than the residual deformation weight. Higher than dynamic and static deformation weights, with a value range of... Fabric wrinkles directly reduce the smoothness of wear and visual quality, so a penalty mechanism with the second highest weight is set to strengthen the deduction of wrinkle defects. The normalized wrinkle deformation rate is the dimensionless value of the original wrinkle deformation rate range, in the range [0,1]. It represents the curvature abrupt deviation caused by local concavity and stacked wrinkles on the partitioned surface, and quantifies the severity of fabric wrinkles. In this embodiment, edge detection algorithms combined with illumination correction and dual-threshold denoising are used to extract the edge features and surface contours of the underwear structure, providing boundary support for region division. Based on the stress structure characteristics of the underwear, partitioning is completed, achieving partition correspondence between the two-dimensional image and the three-dimensional surface model, ensuring the independence of deformation detection. Relying on the principle of two-dimensional pixel geometric registration and the analytical coupling of three-dimensional surface differential curvature, four types of deformation features—static, dynamic, residual, and wrinkled—are distinguished from each other in multiple dimensions, effectively separating reversible elastic deformation from irreversible plastic deformation and quantifying various deformation defects. A comprehensive deformation index is constructed through differentiated weight configuration and range normalization fusion algorithms, highlighting the evaluation weights of core defects such as residual deformation and wrinkled deformation, and unifying the evaluation benchmark for multi-dimensional deformation indicators. This solves the problems of single-dimensional deformation detection, insufficient accuracy, and chaotic evaluation standards in underwear deformation detection, achieving comprehensive, multi-condition, and refined deformation quantification evaluation, and improving the objectivity, completeness, and standardization of underwear deformation detection.

[0020] A fourth embodiment of a method for detecting the degree of deformation of underwear according to the present invention includes: 401. Based on the preset standard fitting elastic rebound attenuation curve and the preset rebound performance calculation formula, analyze the laser profile scanning surface data and segmented fatigue detection data to obtain the rebound performance index. In this embodiment, a pre-defined standard fitting elastic rebound decay curve and a rebound performance calculation formula constitute a long-term elastic rebound quantification dual-layer analytical constraint, used to fuse static deformation information of the laser contour scanning surface and segmented fatigue information for coupled calculation; the overall solution logic is to uniformly quantify the elastic rebound retention capacity of the underwear section after tens of thousands of wear fatigue cycles. The calculation process must simultaneously satisfy two basic analytical constraints: first, the segmented fatigue test data is stored at fixed intervals of 10,000, 20,000, 30,000, 40,000, and 50,000 cycles, and the fluctuating noise data during the test start and stop phases are removed, retaining only samples from the deformation stable interval; second, the laser contour scanning surface data is denoised and segmented. After point completion, the data is divided into sections according to the underwear structure. The surface deformation data of the same section is bound and matched with the fatigue samples of the corresponding grade to avoid cross-regional data mixing and interference with accuracy. First, the paired surface bending deformation and segmented fatigue samples are mapped to the standard elastic rebound attenuation curve for nonlinear fitting and comparison to solve the fatigue attenuation coefficient of each section. Then, the fatigue attenuation coefficient and the section benchmark bending deformation parameters are substituted into the rebound performance calculation formula, and positive normalization correction term and fatigue loss deduction term are introduced to complete the deviation compensation correction. Finally, the dimensionless rebound performance index in the range of 0 to 1 is output, which provides the core evaluation parameters for long-term fatigue resistance deformation for subsequent multi-dimensional coupled quantitative rating. 402. Based on the quantitative rating method, the motion time series data, stress time series data, comprehensive deformation index and rebound performance index are rated and analyzed to obtain the bending deformation level; In this embodiment, the system calls a preset quantitative rating model to conduct multi-dimensional fusion and grading judgment. The overall rating logic is to objectively grade deformation defects by comprehensively considering static deformation, deformation response time difference, deformation hysteresis, and long-term fatigue rebound. The rating process must simultaneously meet two basic judgment constraints: First, the motion time sequence and force time sequence data are pre-timestamped and noise filtered, and the two types of dynamic defect indicators, deformation response time difference and deformation hysteresis, are obtained by cross-correlation feature matching. Second, the comprehensive deformation index, rebound performance index, and dynamic hysteresis parameter are uniformly normalized and mapped to the dimensionless range of 0 to 1 to eliminate the weight imbalance problem caused by the difference in the dimensions of different parameters. Fixed fusion weights are assigned according to the mechanical influence of each parameter on the bending fit quality of underwear. A single defect hard threshold locking correction mechanism is embedded simultaneously. Dynamic hysteresis exceeding the standard, rebound index too low, and comprehensive deformation exceeding the threshold are all subject to graded deduction correction. The corrected comprehensive score is matched with the preset five-level deformation threshold range to output a standardized bending deformation level, thus fully realizing the integrated objective rating of static, dynamic, and long-term fatigue deformation defects. In this embodiment, a quantitative analysis system is constructed by fitting a standard elastic rebound attenuation curve and a rebound performance calculation formula. Coupled analysis is performed on the fatigue data of underwear segments and the deformation data of laser contour surfaces. By using constraints such as fixed fatigue level sampling, eliminating start-stop fluctuation noise, and binding partitioned data, detection errors such as data mixing, sample distortion, and regional misalignment are effectively avoided, ensuring the accuracy and stability of fatigue deformation analysis. The fatigue attenuation coefficient of each partition is solved by nonlinear curve fitting, and deviation compensation correction is completed by combining the benchmark bending deformation parameters. Relying on the dual calculation logic of positive normalization correction and fatigue loss deduction, a standardized rebound performance index in the range of 0 to 1 is output. This solution realizes the quantification of long-term fatigue deformation and elastic rebound loss of underwear, unifies the evaluation benchmarks of different structural partitions and different deformation conditions, makes up for the defects of difficult detection in quantifying long-term anti-relaxation and rebound attenuation performance, and effectively improves the standardization of underwear fatigue rebound quality evaluation.

[0021] A fifth embodiment of a method for detecting the degree of deformation of underwear according to the present invention includes: 501. Based on the preset reference curvature parameters, the preset reference surface, and the preset minimum distance iterative matching algorithm, perform point cloud registration on the underwear surface model to obtain the surface curvature fitting deformation characteristics. In this embodiment, the preset reference curvature parameters, the standard stress-free reference surface, and the minimum distance iterative matching algorithm together constitute the underlying constraints for point cloud registration. These constraints are used to align and compare the 3D surface model of the underwear under inspection conditions with the factory-standard reference surface in the entire spatial domain. The overall registration solution logic eliminates global coordinate deviations caused by tooling clamping offsets and overall fabric displacement, extracting the curvature difference features caused only by fabric deformation. The registration process must simultaneously satisfy two basic constraints: first, the reference surface is the initial stress-free and deformation-free standard 3D model of the underwear, with fixed reference curvature parameters for each partition bound as a curvature comparison scale; second, the minimum distance iterative matching algorithm sets the iteration... The convergence threshold and coordinate error termination condition are used to iteratively reduce the spatial distance between the real-time surface point cloud and the reference surface point cloud until the global coordinate deviation converges to within the preset accuracy. By comparing the global point-to-point spatial deviation, the global rigid body displacement and local flexible deformation are separated, and the surface curvature fitting deformation features corresponding to each detection zone of the underwear are output, providing the original input data of surface deviation for subsequent bending deformation fitting. The basic source model of the minimum distance iterative matching algorithm is the classic ICP (Iterative Closest Point) point cloud registration model, which belongs to the rigid point cloud iterative alignment optimization algorithm based on Euclidean spatial distance metric. 502. Based on the preset curvature fitting algorithm, the surface curvature fitting deformation characteristics and laser contour scanning surface data are fitted and analyzed to obtain the degree of bending deformation. In this embodiment, after point cloud registration and extraction of surface curvature fitting deformation features, the system calls a preset curvature fitting algorithm to integrate high-precision laser contour scanning surface discrete contour data to perform partition curvature analysis calculations. The fitting solution logic quantifies the deformation amplitude of each region of the underwear surface deviating from the standard reference curvature. The calculation process simultaneously satisfies two fitting constraints: First, the laser contour scanning surface data undergoes noise reduction and breakpoint completion preprocessing to completely preserve the continuous surface contours of areas such as cups, side wings, and elastic bands. Second, the surface curvature fitting deformation features are independently separated according to the underwear structure partitions and matched one-to-one with the laser contour data of the same partition to avoid cross-region deformation data mixing and interference with calculation accuracy. The least squares curvature fitting method is used to solve the continuous curvature of the partition surface discrete contours. The difference between the real-time curvature of the partition and the reference curvature parameter is compared to quantify the degree of bending deformation of the standardized partition, providing core parameters of static surface deformation dimension for fatigue attenuation fitting analysis. 503. Analyze the bending deformation degree and segmented fatigue test data according to the standard fitted elastic rebound decay curve and the elastic rebound energy calculation formula to obtain the elastic rebound energy index. In this embodiment, based on the standard fitted elastic rebound decay curve, nonlinear coupling fitting is carried out by binding 10,000 to 50,000 multi-level segmented fatigue steady-state data with the corresponding zone bending deformation degree. Abnormal fluctuation data during the fatigue start-stop stage are eliminated, so that the deformation sample points correspond one-to-one with the fatigue conditions. The zone fatigue decay coefficient, which characterizes the elastic loss of the fabric, is solved. The second level calls the preset rebound performance normalization calculation formula, using the fatigue-free baseline rebound parameter and the zone baseline bending deformation parameter as the correction benchmark. The fatigue decay coefficient is substituted into the calculation as the elastic loss correction variable. At the same time, the zone curvature deviation compensation term is introduced to eliminate the calculation error caused by the original curvature difference of the pattern. Finally, the dimensionless rebound performance index in the range of 0 to 1 is output. The higher the value, the stronger the fabric's fatigue resistance and long-term rebound and shape retention ability. It serves as an independent long-term deformation evaluation dimension to support the subsequent multi-parameter quantitative rating. In this embodiment, the point cloud registration of the underwear surface is completed based on the minimum distance iterative matching algorithm, and the deformation features of the pure curved surface curvature fitting are extracted to avoid the deformation calculation distortion caused by the overall displacement. Then, combined with the preprocessed laser contour data, the degree of bending deformation of each zone is quantified by least square curvature fitting to fully restore the local surface deformation differences of the underwear. Subsequently, multi-level segmented fatigue detection data are linked, and the fatigue decay coefficient is obtained by fitting the standard elastic rebound decay curve. After normalization by a special calculation formula with curvature compensation, a unified rebound performance index in the range of 0 to 1 is output. This solution constructs a layered long-term deformation evaluation link of surface registration, curvature quantification, and fatigue loss analysis. It eliminates multiple interferences such as coordinate offset, contour noise, and original curvature of the pattern layer by layer, unifies the fatigue rebound evaluation benchmark of various underwear, and makes up for the long-term anti-relaxation evaluation dimension missing in static and dynamic detection, thereby improving the accuracy of the long-term deformation quality judgment of underwear and the standardization and comparability of industrial testing.

[0022] The sixth embodiment of a method for detecting the degree of deformation of underwear according to the present invention includes: 601. Based on the standard fitted elastic rebound attenuation curve, perform attenuation analysis on the degree of bending deformation and segmented fatigue test data to obtain the fatigue attenuation coefficient; In this embodiment, the preset standard fitting elastic rebound decay curve incorporates elastic loss benchmark constraints corresponding to multiple fatigue cycles. This constraint is used to perform synchronous coupled decay fitting analysis on segmented fatigue test data and the degree of bending deformation in each zone. The overall solution logic is to quantify the loss of elastic performance of the underwear fabric after multiple cycles of stretching fatigue. The calculation process must simultaneously satisfy two basic fitting constraints: First, the segmented fatigue test data is stored in segments according to fixed fatigue intervals, including multiple sets of steady-state deformation data at 10,000, 20,000, 30,000, 40,000, and 50,000 cycles, excluding abnormal fluctuation data during the start and stop phases of fatigue testing. Second, the degree of bending deformation matches the curvature change of the corresponding zone surface, and is bound one-to-one with the fatigue deformation data of the same zone to ensure a one-to-one correspondence between deformation and fatigue conditions. By mapping multiple sets of fatigue deformation sample points to the standard elastic rebound decay curve to complete nonlinear fitting comparison, the change ratio of fabric elasticity loss with fatigue cycles is solved, and the fatigue decay coefficient corresponding to each test zone of the underwear is output, providing core input parameters for the fatigue loss dimension for subsequent quantitative calculation of the rebound performance index. 602. Calculate the fatigue attenuation coefficient according to the rebound performance calculation formula to obtain the rebound performance index; In this embodiment, after obtaining the fatigue attenuation coefficient of each zone through attenuation fitting analysis, the system retrieves the preset rebound performance calculation formula to perform normalization and quantification calculations. The rebound performance calculation formula has built-in positive normalization correction terms and fatigue loss deduction terms, using fatigue-free baseline rebound parameters as the calculation benchmark and substituting the fatigue attenuation coefficient as the elastic loss correction variable into the formula calculation. During the calculation process, the baseline bending deformation parameters of each zone are simultaneously bound for compensation correction to eliminate calculation deviations caused by differences in initial elasticity in different regions. Finally, a dimensionless standardized rebound performance index in the range of 0 to 1 is output. The higher the index value, the stronger the fabric's fatigue resistance and rebound recovery ability. This rebound performance index can be used as an independent evaluation dimension in addition to static deformation and dynamic time-series parameters, participating in subsequent multi-parameter coupled quantitative rating to achieve a standardized numerical characterization of the underwear's long-term anti-sagging and shape-preserving capabilities. The expression of the rebound performance calculation formula is as follows: , In the formula, To standardize the rebound performance index, it is ultimately constrained to the range of 0 to 1; the first term The first term is a positive normalization correction term, characterizing the inherent resilience of the material; the smaller the fatigue attenuation, the higher the score of this term. The second term... This is a deduction item for fatigue deformation loss, used to deduct the springback deterioration deviation caused by the accumulation of bending deformation; A fixed constant is preset for the weighting coefficient of the positive normalization correction for rebound integrity, which represents the proportion of the contribution of fatigue attenuation to rebound performance, and its value range is [missing information]. Used to magnify / reduce the calculation percentage of the inherent elasticity of the fabric, and to unify the evaluation criteria for underwear of different fabrics and structures; The fatigue attenuation coefficient for each zone is a real-time variable obtained by coupling and fitting the standard fitted elastic rebound attenuation curve with the segmented fatigue test data. It characterizes the elastic loss of the corresponding zone of the underwear after tens of thousands of cycles of tensile fatigue. The larger the value, the more severe the fatigue deterioration of the elastic fibers. The fatigue-free reference attenuation coefficient standard unloaded initial reference constant is the basic attenuation coefficient of the underwear in a brand new, undamaged state without any reciprocating fatigue test, and is used as the unified reference zero point for all fatigue attenuation calculations; It is the maximum attenuation coefficient of ultimate fatigue, the industry-standard critical constant, the ultimate attenuation coefficient corresponding to when the fabric elasticity completely fails and the resilience is completely lost. It is used to normalize and scale the fatigue difference, mapping the fatigue loss term to the 0~1 range. A fixed constant is preset for the weighting factor of bending deformation loss, which characterizes the negative impact of cumulative bending deformation on rebound performance, and its value range is [missing information]. Used to quantify the degree of loss of elastic performance caused by permanent bending deformation of curved surfaces; To determine the real-time bending deformation of the partition, the real-time surface curvature deformation rate is obtained based on laser contour scanning surface data and curvature fitting algorithm, which characterizes the actual surface bending deformation amplitude of the underwear partition under fatigue conditions. The standard curvature deformation rate corresponding to the initial stress-free reference surface of the underwear, used as the reference for bending deformation compensation and correction of each zone, eliminates the calculation deviation caused by the difference in the original pattern curvature of different cups, shoulder straps, and underbust zones. The maximum bending deformation rate is a constant for calibration. It is the curvature deformation rate of the underwear fabric when it reaches the critical state of irreversible permanent bending deformation. It is used to normalize and scale the bending deformation difference. In this embodiment, based on a preset standard fitting elastic rebound decay curve, coupled nonlinear fitting is performed on segmented fatigue test data and zone bending deformation degree. Strict constraints are imposed on the screening of multi-level fatigue steady-state data and the binding of deformation and fatigue conditions. The fatigue decay coefficient of each zone is solved, quantifying the loss of elastic fibers in the fabric under long-term cyclic wear. Then, a benchmark parameter compensation and correction mechanism is introduced through a dedicated rebound performance calculation formula, distinguishing between positive rebound correction and bending loss deduction, and outputting a unified 0-1 dimensionless rebound performance index. This eliminates calculation deviations caused by different patterns and initial elasticity and curvature of different zones. This solution constructs a complete long-term deformation evaluation link through fatigue fitting and normalized quantification, independently supplementing the fatigue resistance evaluation dimension missing in static and dynamic detection. The evaluation benchmark is unified, and the results can be compared horizontally. It can distinguish the differences in long-term anti-sag and shape retention performance of underwear, improving the completeness, standardization, and accuracy of underwear deformation level assessment and service life evaluation.

[0023] The seventh embodiment of a method for detecting the degree of deformation of underwear according to the present invention includes: 701. Based on the preset multi-source time-series cross-correlation feature matching analysis method, perform joint feature analysis on motion time-series data and force time-series data to obtain deformation response time difference and deformation hysteresis. In this embodiment, the pre-defined multi-source temporal cross-correlation feature matching analysis method incorporates three underlying constraints: unified timestamp alignment, temporal phase matching, and peak time series retrieval. These constraints are used to perform coupled feature analysis and deviation quantification calculation on the synchronized motion time series data and force time series data. The overall judgment and solution logic involves calculating the temporal offset between the trigger time of human motion and the peak time of the deformation and force response of the underwear fabric. The calculation process must simultaneously satisfy two basic matching constraints: first, the motion time series and force time series use a globally unified acquisition time reference to complete time series alignment correction, eliminating hard acquisition... The first step is to address the inherent offset error caused by the timing misalignment of the components. The second step is to perform noise filtering preprocessing on the two sets of time-series data to remove invalid fluctuation noise caused by equipment shaking and fabric micro-disturbances, retaining only the effective feature sequences corresponding to the effective movement cycles of human torso rotation, arm raising, flexion and extension, and breathing. By performing cross-correlation comparison of the timing phase offset, peak amplitude difference, and cycle synchronization degree for each movement cycle, the dynamic deformation response feature quantification solution is completed, and the deformation response time difference and deformation hysteresis corresponding to each underwear detection zone are output, providing dynamic deformation dimension input parameters for subsequent multi-parameter coupled quantitative rating. 702. Based on the quantitative rating method, the comprehensive deformation index, deformation response time difference, deformation hysteresis and rebound performance index are rated and analyzed to obtain the deformation degree level. In this embodiment, the system automatically retrieves a preset quantitative rating model, integrates three dimensions of parameters—static deformation, fatigue rebound, and dynamic response—to conduct a graded judgment. The comprehensive deformation index, deformation response time difference, deformation hysteresis, and rebound performance index of different dimensions are first uniformly normalized and mapped to a dimensionless range of 0-1. Then, based on the mechanical contribution of each parameter to the bending deformation quality of the underwear, fixed weight coefficients are assigned to complete weighted fusion. The weight allocation follows the configuration rule that irreversible fatigue damage has a higher priority than static deformation, and static deformation has a higher priority than dynamic following characteristics: rebound performance index weight 0.4, comprehensive deformation index weight 0.3, deformation hysteresis weight 0.15, deformation response time difference weight 0.15, and the sum of the four weights is 1. Simultaneously, a single-defect hard threshold correction logic is embedded. The model execution order follows the hard threshold defect judgment priority > weighted fusion scoring, with defect priority from high to low as rebound failure defects > dynamic bending hysteresis defects. If deformation... If the response time difference exceeds the dynamic synchronization standard threshold, or the deformation lag exceeds the allowable deviation range, the underwear is judged to have a dynamic bending lag defect, and the comprehensive fusion score is downgraded and deducted. If the rebound performance index is lower than the minimum rebound qualified threshold, the bending rebound failure defect is judged to exist, and the rating upper limit is directly locked. The comprehensive score after defect correction is matched one by one with the preset five-level deformation threshold range. The five-level deformation level is ranked as follows: Level 1 (best) > Level 2 (good) > Level 3 (qualified) > Level 4 (obvious defect) > Level 5 (serious defect). The larger the numerical level, the more serious the bending deformation of the underwear and the lower the quality. The standardized deformation degree level is output. This deformation degree level is the final bending deformation level of the underwear. It fully realizes the objective quantitative rating of static deformation accumulation, fatigue rebound attenuation, and dynamic following lag multi-dimensional coupling, avoiding the industry testing drawbacks such as misjudgment, insufficient discrimination, and missed detection of hidden defects caused by single index evaluation. In this embodiment, a multi-source temporal cross-correlation feature matching algorithm is used to couple and analyze motion time-series data and force time-series data. Relying on timestamp alignment, phase matching, and peak retrieval mechanisms, it effectively eliminates temporal misalignment and noise interference, quantifies the dynamic deformation response time difference and deformation hysteresis of underwear, and makes up for the shortcoming of the inability to quantify dynamic following hysteresis defects. At the same time, this solution adopts a multi-parameter normalized weighted fusion combined with a defect threshold locking correction mechanism to link the rebound performance index, comprehensive deformation index, and dynamic deformation parameters for rating, realizing multi-dimensional evaluation of static deformation, fatigue rebound, and dynamic response. It effectively solves the problems of single evaluation dimension, large subjective error, easy omission of hidden defects, and vague grading standards in underwear deformation detection and evaluation, improves the objectivity and repeatability of bending deformation level assessment, realizes refined and standardized quantitative judgment of underwear dynamic deformation defects, and improves the overall reliability and industrial applicability of underwear deformation quality detection.

[0024] Figure 3This is a schematic diagram of the structure of a bra deformation detection device 900 provided in an embodiment of the present invention. The bra deformation detection device 900 can vary significantly due to different configurations or performance. It may include one or more central processing units (CPUs) 910 (e.g., one or more processors) and a memory 920, and one or more storage media 930 (e.g., one or more mass storage devices) storing application programs 933 or data 932. The memory 920 and storage media 930 can be temporary or persistent storage. The program stored in the storage media 930 may include one or more modules (not shown in the diagram), each module may include a series of instruction operations on the bra deformation detection device 900. Furthermore, the processor 910 may be configured to communicate with the storage media 930 and execute a series of instruction operations in the storage media 930 on the bra deformation detection device 900 to implement the steps of the bra deformation detection method provided in the above-described method embodiments.

[0025] An underwear deformation detection device 900 may further include one or more power supplies 940, one or more wired or wireless network interfaces 950, one or more input / output interfaces 960, and / or one or more operating systems 931, such as Windows Server, MacOSX, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 3 The structure of the underwear deformation detection device shown does not constitute a limitation on the underwear deformation detection device. It may include more or fewer parts than shown, or combine certain parts, or have different arrangements of parts.

[0026] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the steps of a method for detecting the degree of deformation of underwear.

[0027] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system, device, or unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0028] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0029] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for detecting the degree of deformation of underwear, characterized in that, include: Collect full-domain high-resolution baseline images and full-domain labeled point datasets of underwear; The surface of the entire domain marked point dataset is fitted using a pre-defined B-spline surface interpolation algorithm to obtain the underwear surface model. Deformation analysis is performed on the full-domain high-definition reference image and the underwear surface model based on the preset reference curvature parameters, the preset deformation index calculation formula and the preset deformation analysis method to obtain the comprehensive deformation index. Acquire laser contour scanning surface data, segmented fatigue detection data, motion timing data, and force timing data of the scanning area; The laser profile scanning surface data, segmented fatigue detection data, motion time series data, stress time series data, and comprehensive deformation index are evaluated and analyzed according to the preset quantitative rating method to obtain the bending deformation level.

2. The method for detecting the degree of deformation of underwear as described in claim 1, characterized in that, The method of performing surface fitting on the global marked point dataset based on a preset B-spline surface interpolation algorithm to obtain an underwear surface model includes: The global marked point dataset is searched and analyzed according to the preset neighborhood search algorithm and the preset radius range to obtain sparse marked point datasets and continuous marked point datasets. The B-spline surface interpolation algorithm is used to fit surfaces to sparse and continuous marker point datasets to obtain the underwear surface model.

3. The method for detecting the degree of deformation of underwear as described in claim 1, characterized in that, The process involves performing deformation analysis on a full-domain high-definition reference image and an underwear surface model based on preset reference curvature parameters, a preset deformation index calculation formula, and a preset deformation analysis method to obtain a comprehensive deformation index, including: Feature extraction is performed on the full-domain high-definition reference image based on the preset edge detection algorithm to obtain edge features and surface contours; The full-domain high-definition reference image is divided into regions based on edge features and surface contours to obtain multiple detection zones; Deformation analysis was performed on multiple detection zones and underwear surface models based on the reference curvature parameters and deformation analysis methods to obtain static deformation rate, dynamic deformation rate, residual deformation rate and wrinkle deformation rate; The static deformation rate, dynamic deformation rate, residual deformation rate, and fold deformation rate are calculated according to the deformation index calculation formula to obtain the comprehensive deformation index.

4. The method for detecting the degree of deformation of underwear as described in claim 1, characterized in that, The process involves performing a rating analysis on laser profile scanning surface data, segmented fatigue detection data, motion time series data, stress time series data, and comprehensive deformation index according to a preset quantitative rating method to obtain the bending deformation level, including: The laser profile scanning surface data and segmented fatigue detection data are analyzed based on the preset standard fitting elastic rebound attenuation curve and the preset rebound performance calculation formula to obtain the rebound performance index. The bending deformation level is obtained by rating and analyzing the motion time series data, stress time series data, comprehensive deformation index and rebound performance index according to the quantitative rating method.

5. The method for detecting the degree of deformation of underwear as described in claim 4, characterized in that, The process involves analyzing laser profile scanning surface data and segmented fatigue detection data based on a preset standard fitted elastic rebound attenuation curve and a preset elastic rebound performance calculation formula to obtain the elastic rebound performance index, including: The point cloud of the underwear surface model is registered based on the preset reference curvature parameters, the preset reference surface, and the preset minimum distance iterative matching algorithm to obtain the surface curvature fitting deformation features. The curvature fitting algorithm is used to fit and analyze the surface curvature fitting deformation characteristics and laser contour scanning surface data to obtain the degree of bending deformation. The bending deformation degree and segmented fatigue test data are analyzed based on the standard fitted elastic rebound decay curve and the elastic rebound performance calculation formula to obtain the elastic rebound performance index.

6. The method for detecting the degree of deformation of underwear as described in claim 5, characterized in that, The bending deformation degree and segmented fatigue test data are analyzed according to the standard fitted elastic rebound decay curve and the elastic rebound energy calculation formula to obtain the elastic rebound energy index, including: The fatigue attenuation coefficient is obtained by performing attenuation analysis on the bending deformation degree and segmented fatigue test data based on the standard fitted elastic rebound attenuation curve. The fatigue attenuation coefficient is calculated based on the rebound performance calculation formula to obtain the rebound performance index.

7. The method for detecting the degree of deformation of underwear as described in claim 4, characterized in that, The process involves evaluating and analyzing motion time-series data, stress time-series data, comprehensive deformation index, and resilience performance index using a quantitative rating method to obtain the bending deformation level, including: Based on the preset multi-source time-series cross-correlation feature matching analysis method, joint feature analysis is performed on motion time-series data and force time-series data to obtain deformation response time difference and deformation hysteresis. The comprehensive deformation index, deformation response time difference, deformation hysteresis, and rebound performance index are rated and analyzed using a quantitative rating method to obtain the deformation degree level.

8. A device for detecting the degree of deformation of underwear, characterized in that, include: The data acquisition module is used to acquire full-domain high-definition baseline images and full-domain marker point datasets of the underwear; The surface fitting module is used to perform surface fitting on the global marked point dataset based on a preset B-spline surface interpolation algorithm to obtain the underwear surface model. The deformation analysis module is used to perform deformation analysis on the full-domain high-definition reference image and the underwear surface model according to the preset reference curvature parameters, the preset deformation index calculation formula and the preset deformation analysis method, so as to obtain the comprehensive deformation index. The data acquisition module is used to acquire laser contour scanning surface data, segmented fatigue detection data, motion timing data, and force timing data of the scanning area; The rating and analysis module is used to perform rating and analysis on laser profile scanning surface data, segmented fatigue detection data, motion time series data, stress time series data and comprehensive deformation index according to a preset quantitative rating method to obtain the bending deformation level.

9. A device for detecting the degree of deformation of underwear, characterized in that, The underwear deformation detection device includes: a memory and at least one processor, wherein the memory stores instructions; At least one of the processors invokes the instructions in the memory to cause the underwear deformation detection device to perform the steps of the underwear deformation detection method as described in any one of claims 1-7.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement the various steps of the underwear deformation detection method as described in any one of claims 1-7.

Images