A sole edge line tracking method based on machine vision

Abstract

The application relates to the technical field of image processing, and particularly discloses a sole edge line tracking method based on machine vision, which is used for solving the problems of large positioning error and low efficiency of accurate tracking of a sole edge contour by manual operation in an existing shoemaking process; three-dimensional contour data of a sole is acquired in real time by a 3D laser scanner as first contour data, and three-dimensional contour data of a standard sole is collected as second contour data; whether the shape of the sole is qualified is evaluated according to first shape features and second shape features of a sole surface, a sole shape difference comparison model is constructed, a two-dimensional image of the sole is acquired by vertical projection on a first sole surface model, a corresponding relationship between adjacent sole edge curves is established, a sole offset between the adjacent sole edge curves is acquired, a sole change trajectory model is further constructed, a sole change trajectory is generated, accurate positioning of sole glue brushing is realized, dynamic correction and deviation early warning are realized, and the stability of a production process is ensured.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, more particularly, the present application relates to a sole edge line tracking method based on machine vision. BACKGROUND

[0002] The shoe production line is a complex automated assembly line, involving various processes such as upper, sole, foot, and last. As a key component of footwear products, the shape, edge profile, and processing precision of the sole directly determine the wearing comfort, durability, and overall appearance of the footwear product. The sole not only needs to meet the ergonomic design requirements, but also needs to consider the fashion and functionality of the product. Therefore, the manufacturing process of the sole puts forward higher requirements on process precision, edge processing, and positioning accuracy. In the traditional shoemaking process, due to limitations in equipment and process means, most of the processes still rely on manual operation. For example, in the placement of the sole, edge line detection, glue application, drying, and buckling processes, manual judgment and operation often have certain errors, and the uncertainty of human operation can easily lead to positioning deviation of the sole in subsequent processes, thereby causing a series of process problems and ultimately affecting the overall quality and production efficiency of the product.

[0003] The placement and edge line processing of the sole mainly rely on manual vision and experience. Workers need to position and detect the sole according to the predetermined template or manual marking on the assembly line. However, in the actual production environment, due to the complex and diverse shape of the sole, slight deformation or deviation may occur during the processing process, combined with the influence of environmental factors such as light and reflection, manual observation often cannot guarantee the consistency and repeatability of precision. Especially in modern shoe factories with high output and high standards, small operational deviations may cause deviations in the work of subsequent automated equipment (such as glue applicators, drying equipment, and buckling devices), thereby causing product defects or scrap phenomena. Therefore, how to reduce the error of manual operation while ensuring efficient production and achieve accurate tracking and positioning of the edge profile of the sole has become one of the technical problems to be solved in the current shoemaking industry. In order to solve the above problems, a technical solution is provided. SUMMARY

[0004] In order to overcome the above-mentioned defects of the prior art, the present application provides a sole edge line tracking method based on machine vision, which generates a sole change trajectory using a sole change trajectory model to achieve accurate positioning of the sole glue brushing, thereby solving the problem of large positioning error and low efficiency in the accurate tracking and positioning of the edge profile of the sole by manual operation in the existing shoemaking process, and solving the problems raised in the background technology.

[0005] To achieve the above-mentioned purposes, the present application provides the following technical solutions:

[0006] A sole edge line tracking method based on machine vision, comprising the following steps:

[0007] Step 1, acquiring three-dimensional profile data of the sole as first profile data in real time through a 3D laser scanner , and selecting a standard sole to collect its three-dimensional profile data as second profile data ;

[0008] Step 2, generating a first sole surface model by processing the first profile data to extract a first shape feature of the sole, generating a second sole surface model by processing the second profile data to extract a second shape feature of the sole, and constructing a sole shape difference comparison model according to the first shape feature and the second shape feature to evaluate whether the sole shape is qualified;

[0009] Step 3, acquiring a two-dimensional image of the sole by vertical projection of the first sole surface model, extracting first outer edge feature data of the sole based on the two-dimensional image of the sole, establishing a corresponding relationship between adjacent sole edge curves by spatiotemporal positioning of the first outer edge feature data, and acquiring a sole offset between the adjacent sole edge curves;

[0010] Step 4, constructing a sole change trajectory model by cumulative summation of the sole offset to generate a sole change trajectory, and realizing precise positioning of sole brushing.

[0011] As a further scheme of the present application, in step 2, according to the first shape feature and the second shape feature, a sole shape difference comparison model is constructed to evaluate whether the sole shape is qualified, and the specific steps are as follows:

[0012] Step 21, filtering the first profile data to obtain , generating a first sole surface model by triangular mesh reconstruction ; extracting a first two-dimensional profile curve by projecting the first sole surface to the horizontal plane , wherein is a normalized parameter representing the proportion of the position on the first two-dimensional profile curve, is the corresponding horizontal coordinate returned to the position, is the corresponding vertical coordinate returned to the position;

[0013] Step 22, filtering the second profile data to obtain , generating a second sole surface model by triangular mesh reconstruction ; extracting a second two-dimensional profile curve by projecting the second sole surface to the horizontal plane , wherein is a normalized parameter representing the proportion of the position on the second two-dimensional profile curve, To return the horizontal coordinates corresponding to this position, To return the vertical coordinates corresponding to that position;

[0014] Step 23, based on the first shoe sole surface model and the second sole surface model Calculate the first similarity measure value; the formula for the first similarity measure value is:

[0015] ;

[0016] In the formula: The first similarity measure value, For the first shoe sole surface model at the h-th sampling point The corresponding height value, For the second shoe sole surface model at the h-th sampling point The corresponding height value, This represents the total number of sampling points;

[0017] Step 24, based on the first two-dimensional contour curve and the second two-dimensional contour curve Calculate the maximum deviation value of the profile curve; the formula for calculating the maximum deviation value of the profile curve is:

[0018] ;

[0019] In the formula: This represents the maximum deviation value of the profile curve. To get the maximum value inside the parentheses, Point p is taken from the first two-dimensional contour curve. The upper boundary, Point q originates from the second two-dimensional contour curve. The lower boundary, Point p originates from the second two-dimensional contour curve The upper boundary, Point q is taken from the first two-dimensional contour curve. The lower boundary, The first two-dimensional contour curve The upper boundary and the second two-dimensional contour curve The maximum distance between the lower boundaries, For the second two-dimensional contour curve The upper boundary and the first two-dimensional contour curve The maximum distance between the lower boundaries;

[0020] Step 25: Based on the first similarity metric and the maximum deviation value of the contour curve, import them into the sole shape difference comparison model to evaluate whether the sole shape is qualified. The formula for the sole shape difference comparison model is:

[0021] ;

[0022] In the formula: This represents the difference in sole shape. The first similarity measure value, This represents the maximum deviation value of the profile curve. The first two-dimensional contour curve With the second two-dimensional contour curve The curvature difference value, The weight of the first similarity measure. The weight of the maximum deviation value of the contour curve. The weights for the curvature difference values;

[0023] Step 26: Compare the sole shape difference value with the preset sole shape difference threshold. If the sole shape difference value is greater than or equal to the preset sole shape difference threshold, the sole shape is unqualified; if the sole shape difference value is less than the preset sole shape difference threshold, the sole shape is qualified.

[0024] As a further aspect of the present invention, in step 3, a two-dimensional image of the sole is obtained by vertically projecting the first sole surface model; the first outer edge feature data of the sole is extracted based on the two-dimensional image; the first outer edge feature data is spatiotemporally located to establish the correspondence between adjacent sole edge curves; and the sole offset between adjacent sole edge curves is obtained. The specific steps are as follows:

[0025] Step 31: Obtain a two-dimensional image of the sole by vertically projecting the first sole surface model. Extracting binary images of shoe sole edges based on image processing Connectivity analysis and morphological operations were used to analyze the binary image of the shoe sole edge. Processing is performed to ensure the continuity of the sole edge lines, and the outline of each sole edge is extracted. Where b is the b-th sole;

[0026] Step 32: Summarize all shoe sole edge lines to obtain the edge contour dataset. ,in, Let B be the b-th sole, and B be the number of sole edge line contours in the edge contour dataset.

[0027] Step 33, for two consecutive shoe sole edge lines and , the matching corresponding relationship between the two shoe sole edge lines is calculated based on feature point matching, and a shoe sole edge line matching error function is:

[0028] ;

[0029] In the formula: is the bth shoe sole edge line, is the b+1th shoe sole edge line, is a normalized parameter, representing the proportion of the position on the b+1th shoe sole edge line;

[0030] In step 4, for the matched corresponding points and , the local offset is:

[0031] ;

[0032] In the formula: is the coordinate of the corresponding point determined by feature point matching on the bth shoe sole edge line , is the coordinate of the corresponding point determined by feature point matching on the b+1th shoe sole edge line .

[0033] As a further scheme of the present application, in step 4, the shoe sole change trajectory model is constructed according to the cumulative summation of the shoe sole offset to generate the shoe sole change trajectory, so as to realize the precise positioning of the shoe sole brushing, and the formula of the shoe sole change trajectory model is:

[0034] ;

[0035] In the formula: is the output of the shoe sole change trajectory model, is the initial local offset, is the local offset of the bth shoe sole edge line, and B is the number of edge contour data sets of the shoe sole edge line contour.

[0036] The technical effect and advantages of the shoe sole edge line tracking method based on machine vision provided by the application are as follows: the 3D laser scanner is used to obtain the three-dimensional profile data of the shoe sole as the first profile data in real time, and the three-dimensional profile data of the standard shoe sole is collected as the second profile data, the shoe sole shape difference comparison model is constructed according to the first shape feature and the second shape feature of the shoe sole surface, and whether the shoe sole shape is qualified is evaluated, the shape feature extraction and standard comparison are realized, the automatic evaluation of the shoe sole qualification is realized, and the manual judgment error is reduced; the first shoe sole surface model is vertically projected to obtain a two-dimensional image of the shoe sole, the corresponding relationship between adjacent shoe sole edge curves is established, the shoe sole offset between adjacent shoe sole edge curves is obtained, accurate basis is provided for the generation of the change trajectory, the shoe sole change trajectory model is further constructed to generate the shoe sole change trajectory, the accurate positioning of the shoe sole glue brushing is realized, so that the glue equipment realizes the accurate positioning, the perfect butt joint of the upper and the shoe sole is ensured, the dynamic correction and deviation early warning are realized, and the stability of the production process is ensured. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 A flowchart of a shoe sole edge line tracking method based on machine vision provided by the application is provided.

[0038] Figure 2 A flowchart of step 2 in the shoe sole edge line tracking method based on machine vision provided by the application is provided.

[0039] Figure 3 A two-dimensional profile comparison analysis diagram provided by the application is provided.

[0040] Figure 4 A shoe sole shape difference thermogram and an index statistical diagram provided by the application are provided.

[0041] Figure 5 A 3D shoe sole surface model diagram provided by the application is provided.

[0042] Figure 6 A shoe sole edge trajectory analysis diagram provided by the application is provided. DETAILED DESCRIPTION

[0043] The technical solutions in the application will be described in detail below with reference to the drawings in the application. Obviously, the described technical solutions are only a part of the application, not all. Based on the technical solutions in the application, all other technical solutions obtained by those skilled in the art without creative labor fall within the scope of protection of the application.

[0044] Example 1

[0045] Figure 1 A flowchart of a shoe sole edge line tracking method based on machine vision provided by the application is provided. AsFigure 1 As shown, a machine vision-based sole edge line tracking method includes the following steps:

[0046] Step 1, real-time acquisition of three-dimensional profile data of the sole by a 3D laser scanner as first profile data , and selection of a standard sole to collect its three-dimensional profile data as second profile data ;

[0047] Step 2, generation of a first sole surface model by processing the first profile data to extract the first shape feature of the sole, generation of a second sole surface model based on the second profile data to extract the second shape feature of the sole, and construction of a sole shape difference comparison model according to the first shape feature and the second shape feature to evaluate whether the sole shape is qualified;

[0048] Step 3, vertical projection of the first sole surface model to obtain a two-dimensional image of the sole, extraction of first outer edge feature data of the sole based on the two-dimensional image of the sole, spatiotemporal positioning of the first outer edge feature data to establish a corresponding relationship between adjacent sole edge curves, and acquisition of a sole offset between adjacent sole edge curves;

[0049] Step 4, construction of a sole change trajectory model according to the cumulative sum of the sole offset to generate a sole change trajectory, and realization of precise positioning of the sole glue brushing.

[0050] The three-dimensional profile data of the sole and the standard sole is collected by a 3D laser scanner, which can accurately capture the slight deformation and details of the sole. The data is processed to construct a sole surface model and extract shape features, and then a shape difference comparison model is established to evaluate whether the sole is qualified. Through shape feature extraction and standard comparison, automatic evaluation of the sole qualification is realized, reducing the error of manual judgment. The first sole surface model is vertically projected to obtain a two-dimensional image, from which the outer edge data of the sole is extracted, and the DTW algorithm is used to realize the spatiotemporal positioning and matching of the continuous edge. The spatiotemporal positioning and matching algorithm enables the subtle offset between the continuous sole edges to be captured, providing accurate basis for change trajectory generation. According to the local offset obtained by matching, a cumulative and smoothing method is used to construct a sole change trajectory model, which provides real-time feedback data for glue positioning. The change trajectory model generated by cumulative sum can dynamically reflect the displacement of the sole in the production process, thereby guiding the precise positioning of the glue equipment to ensure the perfect docking of the upper and the sole. The trajectory data is fed back to the automatic control system to realize dynamic correction and deviation warning, ensuring the stability and consistency of the production process.

[0051] Specifically, as Figure 2The flowchart of step 2 in a machine vision-based sole edge line tracking method is shown, and in step 2, whether the sole shape is qualified is evaluated according to the first shape feature and the second shape feature to construct a sole shape difference contrast model, and the specific steps are as follows:

[0052] Step 21, filtering the first profile data to obtain , and generating a first sole surface model by triangular mesh reconstruction ; extracting the first two-dimensional profile curve by projecting the first sole surface to the horizontal plane , wherein is a normalized parameter representing the proportion of the position on the first two-dimensional profile curve, is the horizontal coordinate corresponding to the position, is the vertical coordinate corresponding to the position;

[0053] Step 22, filtering the second profile data to obtain , and generating a second sole surface model by triangular mesh reconstruction ; extracting the second two-dimensional profile curve by projecting the second sole surface to the horizontal plane , wherein is a normalized parameter representing the proportion of the position on the second two-dimensional profile curve, is the horizontal coordinate corresponding to the position, is the vertical coordinate corresponding to the position;

[0054] Step 23, calculating the first similarity measure value based on the first sole surface model and the second sole surface model ; the formula of the first similarity measure value is as follows:

[0055] ;

[0056] In the formula, is the first similarity measure value, is the height value of the first sole surface model at the hth sampling point , is the height value of the second sole surface model at the hth sampling point , is the total number of sampling points; it should be noted that and represent the sampling points of the two models at the same position, respectively;

[0057] Step 24, calculating the second similarity measure value based on the first two-dimensional profile curve and the second two-dimensional profile curve Calculate the maximum deviation value of the profile curve; the formula for calculating the maximum deviation value of the profile curve is:

[0058] ;

[0059] In the formula: This represents the maximum deviation value of the profile curve. To get the maximum value inside the parentheses, Point p is taken from the first two-dimensional contour curve. The upper boundary, Point q originates from the second two-dimensional contour curve. The lower boundary, Point p originates from the second two-dimensional contour curve The upper boundary, Point q is taken from the first two-dimensional contour curve. The lower boundary, The first two-dimensional contour curve The upper boundary and the second two-dimensional contour curve The maximum distance between the lower boundaries, For the second two-dimensional contour curve The upper boundary and the first two-dimensional contour curve The maximum distance between the lower boundaries;

[0060] Step 25: Based on the first similarity metric and the maximum deviation value of the contour curve, import them into the sole shape difference comparison model to evaluate whether the sole shape is qualified. The formula for the sole shape difference comparison model is:

[0061] ;

[0062] In the formula: This represents the difference in sole shape. The first similarity measure value, This represents the maximum deviation value of the profile curve. The first two-dimensional contour curve With the second two-dimensional contour curve The curvature difference value, The weight of the first similarity measure. The weight of the maximum deviation value of the contour curve. The weights for the curvature difference values;

[0063] Step 26: Compare the sole shape difference value with the preset sole shape difference threshold. If the sole shape difference value is greater than or equal to the preset sole shape difference threshold, the sole shape is unqualified; if the sole shape difference value is less than the preset sole shape difference threshold, the sole shape is qualified.

[0064] likeFigure 5 The two-dimensional contour contrast analysis diagram shown; shows the sole surface model reconstructed by three-dimensional modeling or 3D scanning, which can intuitively see the overall shape, pattern and concave-convex structure of the sole. The length information marked in the figure is used to refer to the size range of the sole under a certain angle, which provides basic data for subsequent shape difference analysis.

[0065] As Figure 4 The shoe sole shape difference heat map and index statistical chart shown; after projecting the sole surface to the two-dimensional plane, two contour curves (usually one is the standard sole contour and the other is the current sole contour) are obtained, and the relative position difference between the green curve and the blue curve intuitively shows the offset and deformation of the two in the X and Y directions. The horizontal axis of the figure is the X value (mm), and the vertical axis is the Y value (mm), which can be used to evaluate the difference between the current sole and the standard contour.

[0066] As Figure 3 The 3D sole surface model diagram shown contains a matrix diagram, which uses color depth or numerical size to represent the difference degree of different sampling points / regions, and gives several shape difference measurement indicators; similarity measurement (d SIM ): the value is 0.892, which represents the root mean square error or similarity of the surface height difference; maximum deviation value (H(C(t),C2(t2))) : 2.45mm, reflecting the maximum offset of the two two-dimensional contour curves at the most extreme position; curvature difference (E cu ): 0.156, used to measure the local bending difference of the two contour curves; total shape difference value (E TOT ): 0.248, which is weighted and synthesized by the above multiple indicators, used to evaluate the closeness of the sole to the standard shape.

[0067] The three-dimensional profile data collected by the 3D laser scanning technology can accurately reflect the micro-deformation of the shoe sole, and the curved surface model obtained by filtering and reconstruction can truly restore the details of the shoe sole surface. The model is compared from two angles: on the one hand, the similarity measurement calculated by the height value (curved surface function) can reflect the overall shape height difference; on the other hand, the maximum deviation and curvature difference of the two-dimensional contour curve can capture the local contour change and detail deviation, and the combination of the two makes the shape comparison more comprehensive; by constructing a comprehensive difference index and introducing a weight coefficient, the influence degree of each index can be flexibly adjusted according to different production requirements, realizing the quantitative evaluation of the shape of the shoe sole, and avoiding the one-sidedness that may be brought by a single index; based on machine vision and mathematical model, the model can automatically collect, process and evaluate the shape data of the shoe sole on the production line, improve the detection speed and consistency, and reduce the error caused by manual intervention; each index (height difference, contour deviation, curvature difference) in the model can be extended and modified according to the actual application requirements, and is suitable for detection of different shoe types and different process requirements, providing a unified platform for intelligent shoemaking; by using multi-dimensional comparison of three-dimensional profile data and two-dimensional contour curve, accurate quantitative evaluation of the shape of the shoe sole is realized, which provides data support for subsequent process control, and has the advantages of high precision, automation and flexible expansion, which can effectively improve the quality and efficiency of the shoemaking production line.

[0068] Specifically, the curvature difference value of the first two-dimensional contour curve and the second two-dimensional contour curve is calculated as follows:

[0069] ;

[0070] In the formula, the curvature difference value of the first two-dimensional contour curve and the second two-dimensional contour curve is , the total arc length of the second two-dimensional contour curve is , the curvature of the first two-dimensional contour curve is , and the curvature of the second two-dimensional contour curve is .

[0071] Specifically, in step 3, the first shoe sole curved surface model is vertically projected to obtain a shoe sole two-dimensional image, the first outer edge feature data of the shoe sole is extracted based on the shoe sole two-dimensional image, the corresponding relationship between adjacent shoe sole edge curves is established by spatiotemporal positioning of the first outer edge feature data, and the shoe sole offset between adjacent shoe sole edge curves is obtained. The specific steps are as follows:

[0072] Step 31, get the 2D image of the sole by vertical projection of the first sole curved surface model , extract the binary image of the sole edge based on image processing , process the binary image of the sole edge by connected domain analysis and morphological operation to ensure the continuity of the sole edge line and extract each sole edge contour line , wherein b is the bth sole;

[0073] Step 32, aggregate all the sole edge lines to obtain the edge contour data set , wherein b is the bth sole, and B is the number of sole edge line contours in the edge contour data set;

[0074] Step 33, for the two continuous sole edge lines and , calculate the matching corresponding relationship between the two sole edge lines based on feature point matching, and the sole edge line matching error function is:

[0075] ;

[0076] In the formula: is the bth sole edge line, is the b+1th sole edge line, is a normalized parameter representing the proportion of the position on the b+1th sole edge line;

[0077] Step 34, for the matched corresponding points and , the local offset is:

[0078] ;

[0079] In the formula: is the coordinate of the corresponding point determined by feature point matching on the bth sole edge line , and is the coordinate of the corresponding point determined by feature point matching on the b+1th sole edge line .

[0080] ​The first shoe sole curved surface model is used to generate a two-dimensional image of the shoe sole by vertical projection, and the first outer edge feature data of the shoe sole is extracted based on image processing technology. Then, the edge contour of each shoe sole is extracted to form an edge contour data set. A feature point matching method is used to establish a corresponding relationship between the edge curves of the continuous soles and calculate the local offset. The three-dimensional shoe sole curved surface is converted into a two-dimensional image by vertical projection. The mature image processing algorithm (such as Canny edge detection and morphological operation) can obtain stable and continuous shoe sole outer edge data, thereby ensuring the accuracy of subsequent matching. The connected component analysis and morphological operation are used to effectively fill the edge broken area, ensuring that the extracted contour curve has good continuity and integrity, and reducing the matching error caused by edge missing. The feature point-based matching method (such as DTW algorithm) is used to perform time and space positioning on the adjacent shoe sole edges, establish an accurate corresponding relationship, and accurately capture the subtle offset of the shoe sole in the production process due to placement, vibration or micro-deformation. By calculating the local offset, the deformation information between the continuous soles can be quantified, providing accurate data support for subsequent generation of shoe sole change trajectory model and dynamic correction. Real-time collection of shoe sole edge data and calculation of offset between adjacent soles provide online correction basis for the gluing process, which helps to realize high-precision and stable automatic production.

[0081] Specifically, in step 4, the shoe sole change trajectory model is constructed according to the cumulative sum of the shoe sole offset, and the shoe sole change trajectory is generated to realize accurate positioning of the shoe sole glue brushing. The formula of the shoe sole change trajectory model is:

[0082] ;

[0083] In the formula: is the output of the shoe sole change trajectory model, is the initial local offset, is the local offset of the bth shoe sole edge, and B is the number of shoe sole edge contours in the edge contour data set.

[0084] As shown in the shoe sole edge trajectory analysis diagram Figure 6 , two edge trajectories are shown as curves over time. The horizontal axis represents time (seconds), and the vertical axis represents offset (millimeters). The blue curve and the green curve represent the motion / deformation of the compared two shoe sole edges (or the same edge in different coordinate directions). Both trajectories show a periodic change similar to a sine wave, indicating that the edge repeats a similar motion pattern within a certain time interval. The blue curve and the green curve are not synchronized in terms of wave peak and wave trough, reflecting the time phase difference between the two edges (or two directions). The numerical value of the wave peak and wave trough can be used to evaluate the amplitude of the edge displacement, thereby judging the offset or deformation degree of the shoe sole in that time period.

[0085] By accumulating the local offset amount between the continuous shoe sole edges, a continuous change trajectory can be formed, thereby reflecting the displacement and deformation of the shoe sole on the production line in real time. The change trajectory provides accurate real-time positioning information for the gluing equipment, so that the glue brushing device can dynamically correct according to the actual position of the shoe sole, and accurate alignment is achieved. The shoe sole may be affected by small placement errors or vibrations during production. Accumulation can capture and accumulate these offset information, thereby compensating for the position during gluing to ensure seamless docking of the upper and the shoe sole. The change trajectory generated by the vectorized local offset amount data can provide more stable position information than single-frame detection, providing high-precision positioning basis for the glue brushing process. The generated shoe sole change trajectory is not only used for real-time positioning, but also serves as a basis for subsequent data analysis. Through trend analysis of the trajectory change, it can be determined whether there is a systematic deviation in the production process, thereby guiding the optimization and adjustment of process parameters. The change trajectory model and the gluing equipment form a closed-loop feedback system, ensuring that the system can automatically adjust when there is a deviation, thereby improving the consistency and quality of the overall product.

[0086] Example 2

[0087] For the above step 3, the first shoe sole curved surface model is vertically projected to obtain a shoe sole two-dimensional image, the first outer edge feature data of the shoe sole is extracted based on the shoe sole two-dimensional image, the corresponding relationship between the adjacent shoe sole edge curves is established by spatiotemporal positioning of the first outer edge feature data, and the shoe sole offset amount between the adjacent shoe sole edge curves is obtained. Now, an example is given to illustrate:

[0088] In the process of the shoe production line, the contour data of three shoe soles is collected, and the curved surface data of the first shoe sole, the second shoe sole and the third shoe sole is generated by 3D scanning. These data is processed by the following steps:

[0089] At this time, the vertical projection of the first shoe sole curved surface model has been completed, and a two-dimensional image is obtained. The first shoe sole edge binary image is extracted by image processing, and connected component analysis and morphological operation are used to ensure the continuity of the edge line. Finally, the edge contour line of each shoe sole is extracted, and the first shoe sole edge contour curve data extracted from the image is given as shown in Table 1:

[0090]

[0091] The edge contour curves of the second and third shoe soles are extracted from their curved surface data using the same method. The data structure of each curve is the same as that of the first shoe sole edge contour line, and the second shoe sole edge contour curve data extracted from the image is given as shown in Table 2:

[0092]

[0093] The third sole edge profile curve data extracted from the image as shown in Table 3:

[0094]

[0095] Based on the above table data, 3 soles are summarized, and the profile curve of each sole is sampled into several points as the basis for subsequent analysis.

[0096] By the method based on feature point matching, the matching error between adjacent sole edge lines is calculated. For sole 1 and sole 2, we find their corresponding points at each position t by feature point matching , calculate the error between each pair of corresponding points. At t=0.5t, the corresponding point of the first sole is , and the corresponding point of the second sole is , then , continue to calculate the matching error for all t corresponding points until a complete error matrix is obtained.

[0097] After matching by feature points, we get the matched corresponding point set and , calculate the local offset of each pair of corresponding points, such as when t=5, the corresponding points of the first sole and the second sole are respectively:

[0098] ;

[0099] ;

[0100] At this time, the local offset is:

[0101] ;

[0102] Based on the above calculation method, the local offset is calculated for each pair of corresponding points, and the entire change trajectory is constructed.

[0103] The embodiment of the application acquires three-dimensional profile data of the shoe sole as first profile data in real time through a 3D laser scanner, selects a standard shoe sole to collect three-dimensional profile data thereof as second profile data, constructs a shoe sole shape difference comparison model according to the first shape feature and the second shape feature of the shoe sole surface to evaluate whether the shoe sole shape is qualified, realizes automatic evaluation of the shoe sole qualification through shape feature extraction and standard comparison, reduces manual judgment error; acquires a two-dimensional image of the shoe sole through vertical projection on the first shoe sole surface model, establishes a corresponding relationship between adjacent shoe sole edge curves, and acquires a shoe sole offset between adjacent shoe sole edge curves, provides an accurate basis for change trajectory generation, further constructs a shoe sole change trajectory model to generate a shoe sole change trajectory to realize accurate positioning of shoe sole glue brushing, thereby guiding the glue bonding equipment to realize accurate positioning, ensuring perfect butt joint of the upper and the shoe sole, realizing dynamic correction and deviation early warning, and guaranteeing stability of the production process.

[0104] The above merely describes the specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0105] Finally, the above merely describes the preferred scheme of the present application, and is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A machine vision-based method for tracking the edge of a shoe sole, characterized in that, Includes the following steps: Step 1: Acquire the three-dimensional contour data of the shoe sole in real time using a 3D laser scanner as the first contour data. The three-dimensional contour data of a standard shoe sole was collected as the second contour data. ; Step 2: Process the first contour data to generate a first sole surface model and extract the first shape features of the sole. Process the second contour data to generate a second sole surface model and extract the second shape features of the sole. Construct a sole shape difference comparison model based on the first and second shape features to evaluate whether the sole shape is acceptable. The specific steps are as follows: Step 21, process the first contour data Filtering process is performed to obtain The first shoe sole surface model is generated by reconstructing a triangular mesh. The first two-dimensional contour curve is extracted from the cross-section of the first shoe sole surface projected onto the horizontal plane. ,in, The normalization parameter represents the proportion of the position on the first two-dimensional contour curve. To return the horizontal coordinates corresponding to this position, To return the vertical coordinates corresponding to that position; Step 22, for the second contour data Filtering process is performed to obtain The second shoe sole surface model is generated by reconstructing a triangular mesh. The second two-dimensional contour curve is extracted from the second sole surface by projecting it onto a horizontal plane. ,in, The normalization parameter represents the proportion of the position on the second two-dimensional contour curve. To return the horizontal coordinates corresponding to this position, To return the vertical coordinates corresponding to that position; Step 23, based on the first shoe sole surface model and the second sole surface model Calculate the first similarity measure value; the formula for the first similarity measure value is: ； In the formula: The first similarity measure value, For the first shoe sole surface model at the h-th sampling point The corresponding height value, For the second shoe sole surface model at the h-th sampling point The corresponding height value, This represents the total number of sampling points; Step 24, based on the first two-dimensional contour curve and the second two-dimensional contour curve Calculate the maximum deviation value of the profile curve; the formula for calculating the maximum deviation value of the profile curve is: ； In the formula: This represents the maximum deviation value of the profile curve. To get the maximum value inside the parentheses, Point p is taken from the first two-dimensional contour curve. The upper boundary, Point q originates from the second two-dimensional contour curve. The lower boundary, Point p originates from the second two-dimensional contour curve The upper boundary, Point q is taken from the first two-dimensional contour curve. The lower boundary, The first two-dimensional contour curve The upper boundary and the second two-dimensional contour curve The maximum distance between the lower boundaries, For the second two-dimensional contour curve The upper boundary and the first two-dimensional contour curve The maximum distance between the lower boundaries; Step 3: Obtain a two-dimensional image of the sole by vertically projecting the first sole surface model; extract the first outer edge feature data of the sole based on the two-dimensional image of the sole; perform spatiotemporal positioning on the first outer edge feature data to establish the correspondence between adjacent sole edge curves; and obtain the sole offset between adjacent sole edge curves. Step 4: Based on the cumulative summation of the sole offset, construct a sole change trajectory model to generate the sole change trajectory, thereby achieving precise positioning of the sole adhesive application.

2. The machine vision-based shoe sole edge tracking method according to claim 1, characterized in that, Step 2 also includes: Step 25: Based on the first similarity metric and the maximum deviation value of the contour curve, import them into the sole shape difference comparison model to evaluate whether the sole shape is qualified. The formula for the sole shape difference comparison model is: ； In the formula: This represents the difference in sole shape. The first similarity measure value, This represents the maximum deviation value of the profile curve. The first two-dimensional contour curve With the second two-dimensional contour curve The curvature difference value, The weight of the first similarity measure. The weight of the maximum deviation value of the contour curve. The weights for the curvature difference values; Step 26: Compare the sole shape difference value with the preset sole shape difference threshold. If the sole shape difference value is greater than or equal to the preset sole shape difference threshold, the sole shape is unqualified; if the sole shape difference value is less than the preset sole shape difference threshold, the sole shape is qualified.

3. The machine vision-based shoe sole edge tracking method according to claim 1, characterized in that, In step 3, a two-dimensional image of the sole is obtained by vertically projecting the first sole surface model. Based on the two-dimensional image, the first outer edge feature data of the sole is extracted. The first outer edge feature data is spatiotemporally located to establish the correspondence between adjacent sole edge curves, and the sole offset between adjacent sole edge curves is obtained. The specific steps are as follows: Step 31: Obtain a two-dimensional image of the sole by vertically projecting the first sole surface model. Extracting binary images of shoe sole edges based on image processing Connectivity analysis and morphological operations were used to analyze the binary image of the shoe sole edge. Processing is performed to ensure the continuity of the sole edge lines, and the outline of each sole edge is extracted. Where b is the b-th sole; Step 32: Summarize all shoe sole edge lines to obtain the edge contour dataset. ,in, Let B be the b-th sole, and B be the number of sole edge line contours in the edge contour dataset. Step 33, for two consecutive shoe sole edge lines and The matching correspondence between the two shoe sole edge lines is calculated using feature point matching. The shoe sole edge line matching error function is: ； In the formula: For the b-th edge of the sole, For the (b+1)th edge line of the sole, The normalization parameter represents the proportion of the position on the (b+1)th edge of the sole. Step 34, for the matched corresponding points and Local offset for: ； In the formula: For the b-th shoe sole edge line The coordinates of the corresponding points are determined through feature point matching. For the (b+1)th edge line of the sole The coordinates of the corresponding points are determined by feature point matching.

4. The machine vision-based shoe sole edge tracking method according to claim 1, characterized in that, In step 4, a sole change trajectory model is constructed by summing the cumulative sole offsets to generate the sole change trajectory, thus achieving precise positioning of the sole adhesive application. The formula for the sole change trajectory model is: ； In the formula: This is the output of the shoe sole change trajectory model. This is the initial local offset. Let B be the local offset of the b-th sole edge line, where B is the number of sole edge line contours in the edge contour dataset.

Images