A sock style auxiliary design system based on three-dimensional sampling

Abstract

The invention relates to the technical field of sock design, and in particular to a sock style auxiliary design system based on three-dimensional sampling. The invention collects a three-dimensional model of a foot and a three-dimensional model of a sock style, extracts a sock style feature line and a foot feature line for similarity matching, and judges whether the sock style and the foot are consistent; generates corresponding first sock feature points and foot feature points based on each feature line, and maps the first sock feature point to the foot feature point to generate a sock mapping path and a second sock feature point; maps the modified mapping path after the second sock feature point is modified back to the sock style three-dimensional model based on the sock mapping path and a mapping relationship matrix, and generates a modified fourth sock feature point; the method can assist in the sock style design and reduce the related problems of socks being unable to be worn; and after the foot style design is modified, it can be mapped back to the original sock style, thereby improving the design accuracy of the sock style.

Description

Technical Field

[0001] The invention relates to the technical field of sock design, and in particular to a sock style auxiliary design system based on three-dimensional sampling. Background Art

[0002] The style of socks is the basis of their appearance and aesthetics. In order to facilitate display, the socks need to be proofed. The traditional proofing method requires knitting and other methods to generate the physical style. However, this method has a long proofing cycle. If the style needs to be modified, it will take even longer, which seriously affects efficiency.

[0003] With the development of computer technology, the simulation proofing method of three-dimensional modeling by computer has achieved better results than the traditional proofing method. Three-dimensional modeling can quickly proof socks compared with traditional methods, and can quickly modify the sock styles with problems. Compared with the traditional sock proofing method, it greatly shortens the proofing cycle and improves efficiency.

[0004] In the prior art, there is a lack of style simulation combining real foot data for sock style design, and the design is only based on the style of the socks themselves before being worn. As a result, the socks may have a pattern effect that is significantly different from the design or become unwearable after actual wear, resulting in reduced design accuracy and affecting aesthetics and wearability.

[0005] Therefore, a sock style auxiliary design system based on three-dimensional sampling is proposed. Summary of the invention

[0006] The purpose of the present invention is to provide a sock style auxiliary design system based on three-dimensional sampling. First, by collecting a three-dimensional model of the foot and a three-dimensional model of the sock style, the sock style feature line and the foot feature line are extracted for similarity matching to determine whether the sock style and the foot are consistent; then, based on each feature line, a corresponding first sock feature point and a foot feature point are generated, and the first sock feature point is mapped to the foot feature point to generate a sock mapping path and a second sock feature point; finally, the modified mapping path of the second sock feature point is mapped back to the sock style three-dimensional model based on the sock mapping path and the mapping relationship matrix to generate a modified fourth sock feature point; this method can assist in the sock style design and reduce the related problems of socks being unable to be worn; at the same time, after the foot style design is modified, it can be mapped back to the original sock style to improve the design accuracy of the sock style.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A sock style auxiliary design system based on three-dimensional sampling includes a foot data acquisition module, a sock data matching module, a sock data simulation wearing module, a sock wearing recording module, a sock style modification and reverse mapping module and a sock style modification module; specifically:

[0009] A foot data collection module collects foot data and establishes a foot three-dimensional model according to the foot data; the foot three-dimensional model includes a foot three-dimensional model and a calf three-dimensional model, and the foot three-dimensional model and the calf three-dimensional model include multiple sizes;

[0010] A sock data matching module, which obtains a sock style three-dimensional model, extracts the sock style three-dimensional model and the foot three-dimensional model, and generates a sock style feature line and a foot feature line;

[0011] Further, a sock three-dimensional coordinate system of the sock style three-dimensional model is obtained; a foot three-dimensional coordinate system of the foot three-dimensional model is obtained; and a coordinate system similarity between the sock three-dimensional coordinate system and the foot three-dimensional coordinate system is calculated;

[0012] Furthermore, the calculation of the coordinate system similarity calculates the coordinate system similarities of the sock three-dimensional coordinate system and the foot three-dimensional coordinate system at multiple different angles, and obtains sock model images of the sock style three-dimensional model and foot model images of the foot three-dimensional model at multiple different angles, and the number of the multiple different angles is not less than 3;

[0013] If the coordinate system similarities are the same; obtaining a sock model image of the sock style three-dimensional model and a foot model image of the foot three-dimensional model;

[0014] Extracting sock style feature lines of the sock model image and foot feature lines of the foot model image according to the DeepLSD model;

[0015] Performing style similarity matching through the sock style feature line and the foot feature line, and issuing a style mismatch warning if no match is found;

[0016] Furthermore, the calculation of the style similarity matching is:

[0017] ;

[0018] in, is the quantitative similarity judgment threshold, is the style similarity result, is the characteristic line slope similarity function, is the similarity function of the number of feature lines, is the feature line length similarity function, is the characteristic line of the foot, is the sock style feature line, , and is the style similarity weight, ;

[0019] If not satisfied , then no Calculate and directly issue a warning of mismatch of the pattern;

[0020] The sock data simulation wearing module generates corresponding first sock sample feature points and foot feature points according to the successfully matched sock style feature lines and foot feature lines;

[0021] Further, the first sock-like feature point and the foot feature point are obtained at the connection point of the line segments of the sock-like feature line and the foot feature line identified by the DeepLSD model;

[0022] Establishing a mapping relationship matrix between the foot feature points and the first sock feature points, and mapping the first sock feature points to the foot feature points to obtain second sock feature points;

[0023] Further, a mapping relationship matrix is ​​established between the foot feature points and the first sock feature points, including a coordinate mapping matrix and a label mapping matrix, wherein the label mapping matrix is ​​a mapping label of the foot feature points and the first sock feature points; the coordinate mapping matrix includes three-dimensional coordinates of the foot feature points and the first sock feature points and associated position information of the label mapping matrix;

[0024] A sock wearing record module, which obtains a sock sample mapping path of the first sock sample feature point and the second sock sample feature point according to the mapping relationship matrix;

[0025] Further, the first sock-like feature points and the foot feature points to be mapped are acquired according to the mapping relationship matrix;

[0026] Initialize the three-dimensional coordinate system of the socks, establish a proportional mapping function between the three-dimensional coordinate system of the socks and the three-dimensional coordinate system of the foot; map the first sock sample feature point onto the three-dimensional foot model according to the proportional mapping function to generate a starting foot coordinate; generate an ending foot coordinate according to the foot feature point;

[0027] The three-dimensional foot model is used as a basic road surface, and a plurality of paths are generated from the starting foot coordinates to the ending foot coordinates through a path planning algorithm, and the path distances are recorded; the path with the maximum path distance is used as the sock pattern mapping path; the sock pattern mapping path also includes: obtaining a maximum extension distance according to the material taken from the three-dimensional sock pattern model, and if the sock pattern mapping path is greater than the maximum extension distance, issuing a style mismatch warning;

[0028] a sock style modification and reverse mapping module, which records the third sock style feature point and the sock style change mapping path and re-matches the mapping relationship matrix if the second sock style feature point changes; maps the third sock style feature point back to the sock style three-dimensional model according to the mapping relationship matrix, the sock style change mapping path and the sock style mapping path to generate a fourth sock style feature point;

[0029] a sock style modification module, adjusting the first sock sample feature point to the fourth sock sample feature point, and generating a new sock style three-dimensional model;

[0030] Further, a change starting point is generated according to the second sock-like feature point, and a change end point is generated according to the third sock-like feature point; the three-dimensional foot model is used as a basic road surface, and a path planning algorithm is used to generate multiple paths from the change starting point to the change end point and record the distances, and the path with the minimum distance is used as the sock-like change mapping path;

[0031] The sock-like mapping path is proportionally mapped according to the sock-like change mapping path to generate a secondary sock-like mapping path; through the secondary sock-like mapping path, the third sock-like feature point is used as an end point to obtain a corresponding secondary change starting point;

[0032] The starting point of the secondary change is used as a pre-reverse mapping point; if the starting point of the secondary change is not on the three-dimensional model of the foot, the shortest distance point from the starting point of the secondary change to the three-dimensional model of the foot is calculated as the pre-reverse mapping point;

[0033] The pre-reflected points are mapped back onto the sock pattern three-dimensional model through a proportional mapping function between the sock three-dimensional coordinate system and the foot three-dimensional coordinate system to generate the fourth sock pattern feature points.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] 1. The present invention extracts sock style image data and foot 3D image data corresponding to multiple angles through a sock style 3D model and a foot 3D model, extracts corresponding foot style feature lines and sock style feature lines based on a DeepLSD model, and matches them based on the number, length and slope of the feature lines; this method can ensure that the designed sock style size can match the corresponding foot size, and the socks can be normally worn on the foot, avoiding the situation where the designed style cannot match the foot size or cannot be worn, and improving the accuracy of sock style design.

[0036] 2. The present invention matches the first sock-like feature point on the sock three-dimensional model and the second sock-like feature point on the foot three-dimensional model through a mapping matrix relationship; a road surface is established with the foot three-dimensional model, the first sock-like feature point is used as the starting point, and the second sock-like feature point is used as the end point, and a sock-like mapping path is generated through the maximum path of a path planning algorithm; through this method, a corresponding path ratio can be generated and recorded during the mapping process, and at the same time, the path planning algorithm can more naturally simulate the stretching process of the socks when worn, and at the same time, the path record can also be modified after the corresponding design is modified, and the data mapping point can be restored according to the sock-like mapping path, which is convenient for subsequent design modifications. While ensuring the freedom of sock style design, the relevant data can further assist the style design and improve the accuracy of the sock style design.

[0037] 3. After the sock style on the foot is modified, the present invention obtains the change mapping path of the third sock feature point after the second sock feature point is modified through the minimum distance of the path planning algorithm, and maps the third sock feature point back to the sock style three-dimensional model in combination with the saved sock mapping path to generate the modified fourth sock feature point; through this method, after the sock style on the foot is modified, the modified sock style when not worn can be automatically generated by storing and calculating the relevant path, which is convenient for modification by relevant personnel, and at the same time, the accuracy of data mapping can also be ensured based on the stored path data, and the three-dimensional model display before and after wearing can further assist designers and improve the accuracy of sock style design. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a schematic diagram of the system flow of the present invention;

[0039] Figure 2 It is a schematic diagram of the feature line similarity matching model process of the present invention;

[0040] Figure 3 It is a structural schematic diagram of the feature line feature recognition layer of the feature line similarity matching model of the present invention. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] The development of computer technology has changed the problems of long proofing time and long communication time span in the existing sock style design. Through three-dimensional modeling, not only can the style of socks be intuitively displayed, but also the style design can be quickly modified, which greatly improves the efficiency; however, the existing technology lacks the display of the style of socks after wearing, which may result in the situation that the pattern after wearing is greatly different from the expected one or the designed socks cannot be worn, affecting the accuracy of the sock style design. Therefore, the present invention provides a sock style auxiliary design system based on three-dimensional sampling, referring to Figure 1 As shown, the technical solution is as follows:

[0043] A foot data collection module collects foot data and establishes a three-dimensional foot model based on the foot data;

[0044] A sock data matching module obtains a sock style three-dimensional model, extracts the sock style three-dimensional model and the foot three-dimensional model, generates sock style feature lines and foot feature lines; performs style similarity matching through the sock style feature lines and the foot feature lines, and issues a style mismatch warning if no match is found;

[0045] The sock data simulation wearing module generates corresponding first sock feature points and foot feature points according to the successfully matched sock style feature lines and foot feature lines, establishes a mapping relationship matrix between the foot feature points and the first sock feature points, and maps the first sock feature points to the foot feature points to obtain second sock feature points;

[0046] A sock wearing record module, which obtains a sock sample mapping path of the first sock sample feature point and the second sock sample feature point according to the mapping relationship matrix;

[0047] a sock style modification and reverse mapping module, which records the third sock style feature point and the sock style change mapping path and re-matches the mapping relationship matrix if the second sock style feature point changes; maps the third sock style feature point back to the sock style three-dimensional model according to the mapping relationship matrix, the sock style change mapping path and the sock style mapping path to generate a fourth sock style feature point;

[0048] The sock style modification module adjusts the first sock sample feature points to the fourth sock sample feature points to generate a new sock style three-dimensional model.

[0049] By obtaining a three-dimensional model of the foot, it can be used as a data measurement standard when simulating the wearing of socks; by obtaining a three-dimensional model of the sock style and a three-dimensional model of the foot, the sock style feature lines and the foot feature lines are generated for similarity matching calculation; this method can make an advance judgment on the size and wearability of the sock style design before wearing; based on the judgment, relevant warnings can be raised to the designer to avoid loopholes such as inability to wear or size mismatch, thereby improving the accuracy of the sock design style;

[0050] The first sock feature point is generated based on the sock style feature line after preliminary warning screening, and the foot feature point is generated based on the foot feature line; in order to facilitate mapping, the relevant information of each mapping point is recorded by establishing a mapping relationship matrix, which is convenient for subsequent mapping work on the one hand, and on the other hand, if the mapping error is found, it can be checked and corrected according to the matrix; the first sock feature point is used as the starting point, and the foot feature point is used as the second sock feature point, that is, the end point, to map and record the sock mapping path for simulated wearing; through this path, data restoration mapping can be performed according to the path in the subsequent modification process, and the path record can also provide a large amount of basic data, which provides a data basis in the subsequent optimization process. Through these basic data, a large amount of data basis can be provided for the sock design style; improve the accuracy of sock design;

[0051] After simulated wearing, if the socks are modified during the wearing process, the modified change mapping path is recorded, the original sock pattern mapping path is modified, and the sock pattern three-dimensional model is restored through the mapping relationship matrix to form the modified content; this method can be based on the recorded path and mapping data after modification, and these basic data can be mapped back to the sock pattern three-dimensional model after the design modification to make corresponding corrections. While ensuring the convenience of modification, the sock pattern data before and after wearing is provided to relevant designers based on a large amount of basic data and calculation processes, so as to reduce the occurrence of design problems and improve the accuracy of sock design.

[0052] Embodiment 1

[0053] For detailed description, this embodiment is further described through the following content, and the data used in this embodiment are all socks:

[0054] A sock style auxiliary design system based on three-dimensional sampling includes a foot data acquisition module, a sock data matching module, a sock data simulation wearing module, a sock wearing recording module, a sock style modification and reverse mapping module and a sock style modification module; specifically:

[0055] A foot data collection module collects foot data and establishes a foot three-dimensional model based on the foot data; the foot three-dimensional model includes a foot three-dimensional model and a calf three-dimensional model, and the foot three-dimensional model and the calf three-dimensional model include multiple sizes; the foot three-dimensional model and the calf three-dimensional model are both collected using the size of a display plastic model. In the specific implementation process, if stockings or pantyhose are considered, a thigh three-dimensional model can also be added to the existing model in combination with relevant technical personnel, and the specific modification can be made according to the specific implementation situation;

[0056] In the actual three-dimensional style design process, the socks may be produced, and the pattern position may be offset or deformed after wearing due to material problems or size problems. This embodiment shows some types of pattern problems, as shown in Table 1:

[0057] Table 1. The three-dimensional model of sock styles and the pattern problems after actual sock production

[0058]

[0059] In the actual design process, some patterns may be designed at the calf position. In order to take this into consideration and prevent the pattern position from being too far away from the actual position after wearing, and to ensure more accuracy in data simulation wearing, a calf model is added for more accurate simulation wearing. At the same time, to ensure that the size of the sock type may be considered during design, multiple sizes are added for design according to the design object, such as multiple children's sizes, multiple adult sizes, etc. This method can cover foot data of various sizes and types, facilitate the subsequent simulation wearing process, and improve the accuracy of the design;

[0060] A sock data matching module, which obtains a sock style three-dimensional model, extracts the sock style three-dimensional model and the foot three-dimensional model, and generates a sock style feature line and a foot feature line;

[0061] Further, a sock three-dimensional coordinate system of the sock style three-dimensional model is obtained; a foot three-dimensional coordinate system of the foot three-dimensional model is obtained; and a coordinate system similarity between the sock three-dimensional coordinate system and the foot three-dimensional coordinate system is calculated;

[0062] Further, the calculation of the coordinate system similarity calculates the coordinate system similarities of the three-dimensional coordinate system of the socks and the three-dimensional coordinate system of the foot at multiple different angles, and obtains sock model images of the three-dimensional model of the sock style and foot model images of the three-dimensional model of the foot at multiple different angles, and the number of the multiple different angles is not less than 3; the multiple different angles are, for example, but not limited to, a front view, a side view, a top view, a bottom view, etc., and specific angles can be added and modified according to specific implementation conditions, but at least include images of three different angles, such as a front view, a side view, and a top view;

[0063] Before simulating wearing, a preliminary judgment must be made on the sock style 3D model and the foot 3D model, that is, whether the currently designed socks are suitable in terms of the size of the sock opening or the sock tube, whether the sock tube is too short to be worn, etc.; therefore, the simulated wearing is preliminarily judged through image data. Since the 3D model involves a large amount of parameter data, if it is directly applied through the sock style 3D model and the foot 3D model, a large amount of data calculation will be performed, consuming a lot of computing resources; and the current judgment is a preliminary judgment. If a large amount of computing resources are consumed, the cost is too high, so image data is used for preliminary analysis to save resources, so image data is used to facilitate data comparison; if only single-sided image data is analyzed, on the one hand, it is not conducive to subsequent mapping work, and it is easy to reduce the accuracy of mapping and the accuracy of simulated wearing. On the other hand, single-sided image data is not convenient for accurately evaluating whether the sock style has size problems; for this reason, 3 different angle faces are used as the minimum number of faces for evaluation, which can more accurately obtain the specific data of the sock style 3D model for analysis, thereby improving the accuracy of analysis and design;

[0064] This embodiment provides a comparison between the image data analysis speed of the trained DeepLSD model and the parameter comparison speed between three-dimensional models. Five sets of data are used in combination with expert experience, and six angles of images, including the front view, rear view, top view, bottom view, left view and right view, are used for parallel operation analysis. The specific experimental results are shown in Table 2:

[0065] Table 2 Comparison of analysis rates of DeepLSD model and 3D model parameters

[0066]

[0067] From the data in Table 2, it can be seen that the analysis rate of the DeepLSD model is much faster than the analysis rate of the three-dimensional model parameters;

[0068] If the coordinate system similarities are the same; obtaining a sock model image of the sock style three-dimensional model and a foot model image of the foot three-dimensional model;

[0069] Extracting the sock style feature lines of the sock model image and the foot feature lines of the foot model image according to the DeepLSD model; performing style similarity matching through the sock style feature lines and the foot feature lines, and issuing a style mismatch warning if no match is found; the matching judgment is obtained based on the similarity matching score, and if the similarity matching score is greater than or equal to the similarity matching threshold, it is judged to be a match, otherwise it is judged to be a mismatch;

[0070] Furthermore, the calculation of the style similarity matching is:

[0071] ;

[0072] in, is the quantitative similarity judgment threshold, is the style similarity result, is the characteristic line slope similarity function, is the similarity function of the number of feature lines, is the feature line length similarity function, is the characteristic line of the foot, is the sock style feature line, , and is the style similarity weight, ; , and The value of is set according to the specific implementation situation; The default score is 90, which can be modified based on the specific implementation and expert experience;

[0073] If not satisfied , then no Calculate and directly issue a warning of mismatch of the pattern;

[0074] During the calculation process, the feature lines of the DeepLSD model are mainly used as the basic data. The feature lines of the three-dimensional model of the sock style and the feature lines of the foot model are calculated for similarity matching. In order to improve the accuracy of the matching, the length, slope and number of the feature lines are evaluated to measure whether the sock style will have size problems, avoid size problems in the sock style design, and improve the accuracy of the design;

[0075] , and The feature line similarity matching model is used for recognition. The similarity matching model includes a feature line preprocessing layer, a feature line feature recognition layer, a feature line similarity calculation layer and a feature line score output layer. Figure 2 As shown;

[0076] The feature line preprocessing layer is used to perform data preprocessing such as noise reduction on the sock style feature lines and the foot feature lines to generate pre-input data;

[0077] The feature line feature recognition layer is used to extract the pre-input data for feature extraction and generate line feature extraction data; the feature line similarity calculation layer performs feature similarity analysis based on the line feature extraction data and generates a similarity analysis result; the feature line score output layer calculates the score based on the similarity analysis result and outputs the result;

[0078] The feature line feature recognition layer is obtained by the Transformer model, the ResNet model and the GRU model; Figure 3 As shown in the figure, the Transformer model includes the Transformer encoding layer and the Transformer decoding layer, which are used to extract feature data and restore and analyze the overall structural features. The ResNet model uses a 4-layer structure to extract shape features, and the GRU model uses a 6-layer structure to extract line change hidden features.

[0079] During the automatic recognition of feature lines, the recognition accuracy of feature lines will affect the calculation process of style similarity matching. If feature lines are missing, the accuracy of the calculation results of similarity matching will be seriously affected. Therefore, this embodiment adopts multiple models for recognition. Table 3 compares the recognition accuracy of feature lines using different models and the model of this embodiment based on the same data. The results are shown in Table 3:

[0080] Table 3 Recognition accuracy of feature lines by this model and different models

[0081]

[0082] From the results in Table 3, it can be seen that the recognition accuracy of the model in this embodiment is high on the feature line;

[0083] The sock data simulation wearing module generates corresponding first sock sample feature points and foot feature points according to the successfully matched sock style feature lines and foot feature lines;

[0084] Further, the first sock feature point and the foot feature point are obtained at the line segment connection of the sock style feature line and the foot feature line identified by the DeepLSD model; in the specific implementation process, in addition to the connection points obtained by the model, other points can also be added based on expert experience as the first sock feature point and the foot feature point for correspondence, and the specific situation can be added and modified according to the implementation content;

[0085] By obtaining the first sock sample feature points and the foot feature points, the relevant data provided by the model characteristics can be fully utilized for further analysis. In addition, feature points have less interference than feature lines and are easier to calculate, which can reduce the complex amount of system calculations and alleviate the data calculation pressure of subsequent analysis. At the same time, the relevant feature points are also convenient for mapping calculations and provide mapping paths, which is convenient for simulating the style of socks after wearing, so as to provide relevant personnel with more intuitive style data, so as to assist designers in improving the accuracy of design.

[0086] Establishing a mapping relationship matrix between the foot feature points and the first sock feature points, and mapping the first sock feature points to the foot feature points to obtain second sock feature points;

[0087] Further, a mapping relationship matrix is ​​established between the foot feature points and the first sock feature points, including a coordinate mapping matrix and a label mapping matrix, wherein the label mapping matrix is ​​a mapping label of the foot feature points and the first sock feature points; the coordinate mapping matrix includes three-dimensional coordinates of the foot feature points and the first sock feature points and associated position information of the label mapping matrix;

[0088] By establishing the mapping relationship matrix coordinates, it is convenient to record the corresponding point coordinate mapping calculation during mapping. At the same time, by establishing labels for association, it is also convenient to quickly correspond to the relevant points and relevant coordinates based on these labels if there is a problem with the system or mapping work, and to find out whether it is a data error or a calculation error, thereby improving the inspection efficiency. At the same time, it is also convenient to perform reverse mapping based on the relevant data after the three-dimensional model of the sock style is modified in the three-dimensional model of the foot, so as to provide more direct data to the relevant designers, so as to reduce the occurrence of design problems and improve the accuracy of the design from the data level;

[0089] A sock wearing record module, which obtains a sock sample mapping path of the first sock sample feature point and the second sock sample feature point according to the mapping relationship matrix;

[0090] Further, the first sock feature points and the foot feature points to be mapped are obtained according to the mapping relationship matrix; the sock three-dimensional coordinate system is initialized, and a proportional mapping function of the sock three-dimensional coordinate system and the foot three-dimensional coordinate system is established; the first sock feature points are mapped to the foot three-dimensional model according to the proportional mapping function to generate starting foot coordinates; according to the foot feature points, ending foot coordinates are generated; the foot three-dimensional model is used as a basic road surface, and a path planning algorithm is used to generate multiple paths from the starting foot coordinates to the ending foot coordinates, and the path distances are recorded; the path with the maximum path distance is used as the sock mapping path;

[0091] The path planning calculation is performed with the foot as the road surface. According to this method, the stretching of the socks when worn can be simulated based on the existing technology. This method allows designers to more intuitively understand the changes in the graphics of the socks after stretching during wearing, which is convenient for data correction to reduce the occurrence of design problems. At the same time, based on the path planning algorithm, some more complex path data can also be recorded, which is convenient for subsequent reverse mapping work and improves the accuracy of the design from the data calculation level of the mapping work.

[0092] The sock pattern mapping path also includes: obtaining a maximum extension distance according to the material adopted by the sock pattern three-dimensional model, and issuing a style mismatch warning if the sock pattern mapping path is greater than the maximum extension distance; due to the material extension characteristics of the socks, excessive stretching is likely to occur during the stretching process, causing the sock material to break and cause holes; through the path data of the simulation data, the current path distance can be analyzed to simulate the stretching distance, and when a severe stretching situation occurs, an accurate warning can be issued through the simulation data, and the path method combined with the actual application level can further assist relevant designers to reduce the occurrence of design problems and improve the accuracy of the design;

[0093] Embodiment 2

[0094] a sock style modification and reverse mapping module, which records the third sock style feature point and the sock style change mapping path and re-matches the mapping relationship matrix if the second sock style feature point changes; maps the third sock style feature point back to the sock style three-dimensional model according to the mapping relationship matrix, the sock style change mapping path and the sock style mapping path to generate a fourth sock style feature point;

[0095] a sock style modification module, adjusting the first sock sample feature point to the fourth sock sample feature point, and generating a new sock style three-dimensional model;

[0096] Further, a change starting point is generated according to the second sock-like feature point, and a change end point is generated according to the third sock-like feature point; the three-dimensional foot model is used as a basic road surface, and a path planning algorithm is used to generate multiple paths from the change starting point to the change end point and record the distances, and the path with the minimum distance is used as the sock-like change mapping path;

[0097] The sock-like mapping path is proportionally mapped according to the sock-like change mapping path to generate a secondary sock-like mapping path; through the secondary sock-like mapping path, the third sock-like feature point is used as an end point to obtain a corresponding secondary change starting point;

[0098] The starting point of the secondary change is used as a pre-reverse mapping point; if the starting point of the secondary change is not on the three-dimensional model of the foot, the shortest distance point from the starting point of the secondary change to the three-dimensional model of the foot is calculated as the pre-reverse mapping point;

[0099] Mapping the pre-reverse mapping point back to the sock pattern three-dimensional model through a proportional mapping function of the sock three-dimensional coordinate system and the foot three-dimensional coordinate system to generate the fourth sock pattern feature point;

[0100] Specifically, in the mapping process, the mapped pattern may be mapped to the outer or inner layer of the sock 3D model due to calculation problems, but not to the sock style 3D model. Therefore, the shortest distance from each feature point on the pattern to the sock style 3D model is calculated, and all points are mapped to the sock style 3D model using the shortest distance as the reference coordinate.

[0101] In order to evaluate whether the modified pattern after the simulated wearing is modified by the sock style modification module and then mapped back to the sock style 3D model, can the modified pattern be accurately mapped back to the corresponding position on the sock style 3D model; for this purpose, this example adopts a simulation experiment, collects historical sock sample 3D model data with position problems due to the pattern of the foot and its modified historical sock style 3D model data, and the sock style 3D model data are all short socks; through the system of this embodiment, after mapping to the foot and letting relevant technicians make modifications, it maps back to the sock style 3D model data again, and compares the position accuracy with the real modified historical sock sample 3D model data, and the results are shown in Table 4:

[0102] Table 4 Comparison of the accuracy of real modified data and modified data after mapping

[0103]

[0104] It can be seen from Table 4 that the accuracy of the mapped and actual modified positions is about 94%. The higher the accuracy, the better the mapping effect, which can assist designers in designing and improve the accuracy of the design.

[0105] By establishing a mapping matrix, change mapping path and sock sample mapping path, after the relevant technicians modify the sock style on the three-dimensional foot model, they can restore and map it to the three-dimensional sock model based on the recorded data through the changes in relevant feature points. These recorded data not only provide a detailed data basis for review and observation, but also can be used for restoration calculations and provide relevant designers with sock style data before and after wearing, so as to reduce the occurrence of design problems and further improve the accuracy of the design from the perspective of the reusability and reliability of data and data calculation.

[0106] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A sock style auxiliary design system based on three-dimensional sampling, characterized in that: include: A foot data collection module collects foot data and establishes a three-dimensional foot model based on the foot data; A sock data matching module obtains a sock style three-dimensional model, extracts the sock style three-dimensional model and the foot three-dimensional model, generates sock style feature lines and foot feature lines; performs style similarity matching through the sock style feature lines and the foot feature lines, and issues a style mismatch warning if no match is found; The sock data simulation wearing module generates corresponding first sock feature points and foot feature points according to the successfully matched sock style feature lines and foot feature lines, establishes a mapping relationship matrix between the foot feature points and the first sock feature points, and maps the first sock feature points to the foot feature points to obtain second sock feature points; A sock wearing record module, which obtains a sock sample mapping path of the first sock sample feature point and the second sock sample feature point according to the mapping relationship matrix; a sock style modification and reverse mapping module, which records the third sock style feature point and the sock style change mapping path and re-matches the mapping relationship matrix if the second sock style feature point changes; According to the mapping relationship matrix, the sock pattern change mapping path and the sock pattern mapping path, mapping the third sock pattern feature point back to the sock pattern three-dimensional model to generate a fourth sock pattern feature point; A change starting point is generated according to the second sock-like feature point, and a change end point is generated according to the third sock-like feature point; Using the three-dimensional foot model as a basic road surface, a path planning algorithm is used to generate multiple paths from the change starting point to the change end point and the distances are recorded, and the path with the shortest distance is used as the sock-like change mapping path; The sock-like mapping path is proportionally mapped according to the sock-like change mapping path to generate a secondary sock-like mapping path; through the secondary sock-like mapping path, the third sock-like feature point is used as an end point to obtain a corresponding secondary change starting point; The starting point of the secondary change is used as a pre-reverse mapping point; if the starting point of the secondary change is not on the three-dimensional model of the foot, the shortest distance point from the starting point of the secondary change to the three-dimensional model of the foot is calculated as the pre-reverse mapping point; Mapping the pre-reflected points back onto the sock pattern three-dimensional model through a proportional mapping function of the sock three-dimensional coordinate system and the foot three-dimensional coordinate system to generate the fourth sock pattern feature points; The sock style modification module adjusts the first sock sample feature points to the fourth sock sample feature points to generate a new sock style three-dimensional model.

2. The sock style auxiliary design system based on three-dimensional sampling according to claim 1, characterized in that: The foot three-dimensional model includes a foot three-dimensional model and a calf three-dimensional model, and the foot three-dimensional model and the calf three-dimensional model include multiple sizes.

3. The sock style auxiliary design system based on three-dimensional sampling according to claim 1, characterized in that: Extracting the sock style three-dimensional model and the foot three-dimensional model, and generating sock style feature lines and foot feature lines comprises: Acquire a sock three-dimensional coordinate system of the sock style three-dimensional model; acquire a foot three-dimensional coordinate system of the foot three-dimensional model; calculate the coordinate system similarity between the sock three-dimensional coordinate system and the foot three-dimensional coordinate system; if the coordinate system similarities are the same; acquire a sock model image of the sock style three-dimensional model and a foot model image of the foot three-dimensional model; The sock style feature lines of the sock model image and the foot feature lines of the foot model image are extracted according to the DeepLSD model.

4. The sock style auxiliary design system based on three-dimensional sampling according to claim 3, characterized in that: Calculate the coordinate system similarity between the three-dimensional coordinate system of the socks and the three-dimensional coordinate system of the foot at multiple different angles, and obtain sock model images of the three-dimensional model of the sock style and foot model images of the three-dimensional model of the foot at multiple different angles, and the number of the multiple different angles is not less than 3.

5. The sock style auxiliary design system based on three-dimensional sampling according to claim 1, characterized in that: Performing pattern similarity matching using the sock pattern feature line and the foot feature line includes: in, is the quantitative similarity judgment threshold, is the style similarity result, is the characteristic line slope similarity function, is the similarity function of the number of feature lines, is the feature line length similarity function, is the foot feature line, is the sock style feature line, , and is the style similarity weight, If not satisfied , then no Calculation directly issues a warning of style mismatch.

6. The sock style auxiliary design system based on three-dimensional sampling according to claim 1, characterized in that: Generating corresponding first sock feature points and foot feature points based on the successfully matched sock style feature lines and foot feature lines includes: obtaining the first sock feature points and the foot feature points at the connection points of the line segments of the sock style feature lines and the foot feature lines identified by the DeepLSD model.

7. The sock style auxiliary design system based on three-dimensional sampling according to claim 1, characterized in that: Establishing a mapping relationship matrix between the foot feature points and the first sock feature points includes a coordinate mapping matrix and a label mapping matrix, wherein the label mapping matrix is ​​a mapping label of the foot feature points and the first sock feature points; the coordinate mapping matrix includes the three-dimensional coordinates of the foot feature points and the first sock feature points and the associated position information of the label mapping matrix.

8. The sock style auxiliary design system based on three-dimensional sampling according to claim 1, characterized in that: The sock sample mapping path for acquiring the first sock sample feature point and the second sock sample feature point includes: Acquire the first sock-like feature points and the foot feature points that need to be mapped according to the mapping relationship matrix; Initialize the three-dimensional coordinate system of the socks, establish a proportional mapping function between the three-dimensional coordinate system of the socks and the three-dimensional coordinate system of the foot; map the first sock sample feature point onto the three-dimensional foot model according to the proportional mapping function to generate a starting foot coordinate; generate an ending foot coordinate according to the foot feature point; The three-dimensional foot model is used as a basic road surface, and a path planning algorithm is used to generate multiple paths from the starting foot coordinates to the ending foot coordinates, and the path distances are recorded; the path with the maximum path distance is used as the sock-like mapping path.

9. The sock style auxiliary design system based on three-dimensional sampling according to claim 8, characterized in that: The sock pattern mapping path further includes: obtaining a maximum extension distance according to the material adopted by the sock pattern three-dimensional model, and issuing a style mismatch warning if the sock pattern mapping path is greater than the maximum extension distance.

Images