Method for rapid detection of fracture region in precision-stamped part of new energy vehicle

Abstract

The present invention relates to the technical field of image detection, and in particular to a method for rapid detection of a fracture region in a precision-stamped part of a new energy vehicle. The method comprises: acquiring a complete stamped part grayscale image; acquiring each sub-region in the complete stamped part grayscale image; for each sub-region, taking a fitting curve of all edge pixels of the sub-region as an edge curve of the sub-region, and constructing a curvature evaluation factor between each pixel and the adjacent previous pixel on the edge curve; constructing the degree of curvature change of the edge curve; constructing a curvature sequence and a distance sequence of the edge curve; acquiring each subsequence of the curvature sequence of the edge curve; constructing the fracture edge matching degree, a local stress discontinuity index and a fracture region confidence level of the sub-region; and on the basis of the fracture region confidence levels of all sub-regions, using a clustering algorithm to cluster and delineate a fracture region. According to the present invention, the accuracy of detection of the fracture region in the precision-stamped part of the new energy vehicle can be improved.

Description

A new energy vehicle stamping precision part fracture area rapid detection method TECHNICAL FIELD

[0001] The present application relates to the technical field of image detection, in particular to a new energy vehicle stamping precision part fracture area rapid detection method. BACKGROUND

[0002] Stamping is an important metal sheet forming process, which has the advantages of high production efficiency, good interchangeability of parts, low cost, etc. However, in the production process of stamping parts, due to mold wear, matching degree decline, inconsistent falling speed of punch, metal density of sheet metal and other reasons, stamping parts are prone to cracking, scratches and patches. The stamping precision parts of new energy vehicles refer to precision stamping metal parts used to manufacture key components such as vehicle body, chassis and battery pack during the manufacturing process of new energy vehicles. These parts usually require high precision and high strength to meet the requirements of lightweight, safety and energy efficiency of new energy vehicles. With the rapid development of the new energy vehicle industry, stamping precision parts play an increasingly important role in this field. Therefore, it is particularly important to detect the quality of new energy vehicle stamping precision parts.

[0003] At present, the outer contour defect detection of stamping parts includes detection tool detection and image detection methods. The detection tool detection method is easily affected by tools and environment, has high error detection rate, low efficiency and poor flexibility, and has certain limitations. Most of the traditional image detection methods are shallow extraction of defect features, and the detailed feature information is difficult to be utilized, which leads to poor generalization ability. For obvious fracture and easy classification defects, it has high detection accuracy, but for complex shape and similar fracture stamping part edges, it is easy to have false detection problems, and the accuracy is low. SUMMARY

[0004] In order to solve the above technical problems, the present application provides a new energy vehicle stamping precision part fracture area rapid detection method to solve the existing problems.

[0005] The new energy vehicle stamping precision part fracture area rapid detection method of the present application adopts the following technical scheme:

[0006] An embodiment of the present application provides a new energy vehicle stamping precision part fracture area rapid detection method, which comprises the following steps:

[0007] Collecting stamping precision part images, and performing gray scale processing on the stamping precision part images to obtain complete stamping part gray scale images;

[0008] The watershed segmentation algorithm is used to obtain each sub-region in the grayscale image of the complete stamping part, and the edge detection algorithm is used to detect the edge of the sub-region. The fitting curve of all edge pixel points of the sub-region is used as the edge curve of the sub-region. The curvature evaluation factor between each pixel point on the edge curve and the adjacent previous pixel point is constructed according to the neighborhood pixel points of each pixel point on the edge curve; the curvature mutation degree of the edge curve is constructed according to the curvature evaluation factor between each pixel point on the edge curve and the adjacent previous pixel point; the curvature sequence of the edge curve is constructed according to the curvature evaluation factor between each pixel point on the edge curve and the adjacent previous pixel point; according to the coordinates of the pixel points on the edge curve, the curvature sequence of the edge curve is constructed. Construct a distance sequence of the edge curve; obtain each subsequence of the curvature sequence of the edge curve; construct the fracture edge fit of the sub-region based on the curvature sequence of the edge curve, the distance sequence, and each subsequence of the curvature sequence; construct a square window with each pixel point on the edge curve of the sub-region as the center, calculate the MLBP value of the center of the square window using the MLBP algorithm, and construct the local stress discontinuity index of the sub-region based on the grayscale value, MLBP value, and curvature evaluation factor of the pixel point on the edge curve of the sub-region; construct the fracture region confidence of the sub-region based on the local stress discontinuity index of the sub-region, the fracture edge fit, and the curvature mutation degree of the edge curve of the sub-region;

[0009] According to the fracture area confidence of all sub-areas, clustering algorithm is used to cluster and divide the fracture areas.

[0010] Furthermore, the curvature evaluation factor between each pixel point on the edge curve and the adjacent previous pixel point is constructed based on the neighboring pixel points of each pixel point on the edge curve, including: using the least squares method to calculate the tangent at each pixel point on the edge curve, constructing a square window with each pixel point on the edge curve as the center, and the curvature evaluation factor between the pixel point i on the edge curve and the adjacent previous pixel point i-1 The expression is: Where, is the tangent line at pixel i The tangent line to the adjacent previous pixel i-1 The angle between In order to tangent and tangent Convert the angle between them into radians. is the Euclidean distance between pixel i and the adjacent previous pixel i-1, is the number of corner points in the square window centered at pixel i, is the number of corner points in the square window centered at pixel i-1.

[0011] Further, the constructing the curvature mutation degree of the edge curve according to the curvature evaluation factor between each pixel point on the edge curve and the adjacent previous pixel point comprises:

[0012] calculating the mean value of the curvature evaluation factor between all the pixel points except the first pixel point on the edge curve and the adjacent previous pixel point, calculating the absolute value of the difference between the curvature evaluation factor between each pixel point except the first pixel point on the edge curve and the adjacent previous pixel point and the mean value, and taking the mean value of the absolute value of the difference between all the pixel points except the first pixel point on the edge curve and the adjacent previous pixel point as the curvature mutation degree of the edge curve.

[0013] Further, the constructing the curvature sequence of the edge curve according to the curvature evaluation factor between each pixel point on the edge curve and the adjacent previous pixel point comprises: composing the curvature sequence of the edge curve by the curvature evaluation factor between all the pixel points except the first pixel point on the edge curve and the adjacent previous pixel point.

[0014] Further, the constructing the distance sequence of the edge curve according to the coordinates of the pixel points on the edge curve comprises: composing the distance sequence of the edge curve by the Euclidean distance between all the pixel points except the first pixel point on the edge curve and the adjacent previous pixel point.

[0015] Further, the obtaining each subsequence of the curvature sequence of the edge curve comprises: dividing the curvature sequence into subsequences by using a preset length sliding window, wherein the sliding step length of the sliding window is a preset step length.

[0016] Further, the constructing the fracture edge fitting degree of the subregion according to the curvature sequence, the distance sequence and each subsequence of the curvature sequence of the edge curve comprises:

[0017] calculating the absolute value of the Pearson correlation coefficient between the curvature sequence and the distance sequence of the edge curve, calculating the mean value of the Pearson correlation coefficient between the first subsequence and all the other subsequences of the curvature sequence, and taking the product of the absolute value and the mean value as the fracture edge fitting degree of the subregion.

[0018] Further, the constructing the local stress discontinuity index of the subregion comprises:

[0019] Calculate the mean of the grayscale values ​​of all pixels on the edge curve of the sub-area, calculate the absolute value of the difference between the grayscale value of each pixel except the first pixel on the edge curve of the sub-area and the mean, calculate the MLBP value of each pixel except the first pixel on the edge curve plus the curvature evaluation factor between each pixel and the adjacent previous pixel, calculate the product of the absolute value of the difference of each pixel except the first pixel on the edge curve of the sub-area and the calculation result, and calculate the mean of the products of all pixel points on the edge curve of the sub-area except the first pixel as the local stress discontinuity index of the sub-area.

[0020] Furthermore, the constructing of the fracture region confidence of the sub-region includes:

[0021] Calculate the sum of the curvature mutation degree of the edge curve of the sub-region and the local stress discontinuity index of the sub-region, which is recorded as the first sum; calculate the sum of the fracture edge fit of the sub-region and the preset parameter adjustment coefficient greater than zero, which is recorded as the second sum; and use the ratio of the first sum to the second sum as the fracture area confidence of the sub-region.

[0022] Furthermore, clustering and dividing the fractured regions using a clustering algorithm according to the fractured region confidence of all sub-regions includes:

[0023] The fracture area confidence of all sub-areas in the grayscale image of the complete stamping part is used as input, and a clustering algorithm is used to obtain a preset number of clusters. The mean of the fracture area confidence of each cluster is calculated. When the normalized value of the mean of the cluster is greater than the preset fracture area confidence threshold, the sub-area corresponding to the cluster is determined as a fracture area; otherwise, the sub-area corresponding to the cluster is determined as a non-fracture area.

[0024] The present invention has at least the following beneficial effects:

[0025] The present application firstly divides the stamping part into multiple sub-regions, which facilitates better identification of details inside each sub-region; analyzes the bending degree of each place on the edge curve of the sub-region, and calculates the difference in bending degree of each place on the edge curve according to the analysis result, to construct the curvature mutation degree of the edge curve, which takes into account the irregular characteristics of the bending degree of each place on the edge curve of the fracture region; constructs the fracture edge fitting degree of the sub-region according to the correlation degree of the curvature change and the Euclidean distance change at adjacent pixel points on the edge curve of the sub-region, and the correlation degree of the curvature change at each pixel point on the edge curve of the sub-region, which takes into account the regularity of the curvature distribution at each pixel point on the edge curve, and the regularity of the correlation degree between the curvature change and the Euclidean distance change; further combines the uniformity of the edge stress distribution of the sub-region and the tensile texture feature to construct the local stress discontinuity index of the sub-region, to analyze the possibility of the sub-region being a fracture region, and to avoid the problem of misdetecting the edge of the stamping part as the edge of the fracture region; constructs the fracture region confidence of the sub-region according to the curvature mutation degree of the edge curve of the sub-region, the local stress discontinuity index of the sub-region, and the fracture edge fitting degree, to analyze the possibility of the sub-region being a fracture region from multiple aspects, and to improve the accuracy of the analysis of the sub-region being a fracture region; uses a clustering algorithm to cluster and divide the fracture region according to the fracture region confidence of the sub-region, to cluster the sub-regions with similar features into the same clustering cluster according to the fracture region confidence of all sub-regions, so as to further analyze and judge which sub-regions are fracture regions. The present application can improve the accuracy of the fracture region detection of the stamping precision part of the new energy vehicle. BRIEF DESCRIPTION OF DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, and the advantages thereof, a brief introduction will be given to the drawings needed in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor based on these drawings.

[0027] Fig. 1 is a step flow chart of a new energy vehicle stamping precision part fracture region rapid detection method provided by the present application;

[0028] Fig. 2 is a fracture region confidence acquisition flow chart. DETAILED DESCRIPTION

[0029] In order to further clarify the technical means and effects taken by the present application to achieve the predetermined inventive objectives, the following describes in detail the specific implementation, structure, features and effects of a new energy vehicle stamping precision part fracture area rapid detection method according to the present application, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0030] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0031] The specific scheme of the new energy vehicle stamping precision part fracture area rapid detection method provided by the present application is described in detail below in combination with the accompanying drawings.

[0032] One embodiment of the present application provides a new energy vehicle stamping precision part fracture area rapid detection method, specifically, a new energy vehicle stamping precision part fracture area rapid detection method is provided as follows, please refer to FIG. 1, the method comprises the following steps:

[0033] Step S001, the stamping precision part of the new energy vehicle is photographed by the CCD camera, and the collected image is preprocessed.

[0034] The CCD camera is used to photograph the stamping precision part image of the new energy vehicle, and the acquired image is preprocessed to eliminate the influence caused by noise and part of external interference, so as to enhance the accuracy of subsequent analysis. In order to remove noise while retaining boundary information, the median filtering method is selected in this embodiment to process the image, and other denoising methods can also be adopted. Then the acquired RGB image is converted into a gray image, which is denoted as a complete stamping part gray image A. The median filtering method is a known technology, and this embodiment will not be described again.

[0035] Step S002, using a watershed segmentation algorithm to segment the complete stamping part gray image, calculating the curvature mutation degree of the edge curve of the sub-region, constructing the fracture edge fitness, local stress discontinuity index and fracture area confidence of the sub-region.

[0036] The complete stamping part gray image A is segmented using a watershed segmentation algorithm, and is segmented into t sub-regions, so that the details inside the region can be better identified. In this embodiment, t is 100, Canny edge detection is performed on the 100 sub-regions, and Harris corner detection algorithm is used for corner detection. The watershed segmentation algorithm, Canny edge detection algorithm and Harris corner detection algorithm are all known technologies, and this embodiment will not be described again.

[0037] Curvature can be used to describe the degree of bending change of an object. The larger the curvature, the greater the degree of bending of the object. Because the edge contour of the fracture area is forcibly broken by external force, the curvature of the edge contour of the fracture area varies greatly.

[0038] Taking sub-region r as an example, the least squares method is used to perform curve fitting on all edge pixel points of the sub-region, and the above fitting curve is recorded as the edge curve of sub-region r. With each pixel point on the edge curve as the center, a window of L1*L1 is constructed. The value of L1 can be set by the implementer. In this embodiment, L1 is 3. If the edges of the stamping parts initially set are all large changes in curvature, errors may occur if only the curvature calculation is used. Even the edges of good stamping parts with large changes in curvature are smooth and have fewer corners. The tangent line at each pixel point on the edge curve is calculated separately using the least squares method. The least squares method is a well-known technology and will not be repeated in this embodiment. The expression for calculating the curvature evaluation factor is: Where, is the curvature evaluation factor between pixel i and the adjacent previous pixel i-1 on the edge curve, is the tangent line at pixel i The tangent line to the adjacent previous pixel i-1 The angle between In order to tangent and tangent The angle between them is converted into radians and can be calculated with other coefficients. is the Euclidean distance between pixel i and the adjacent previous pixel i-1, is the number of corner points in the window centered at pixel i, is the number of corner points in the window centered at pixel i-1.

[0039] When the edge curve between pixel i and pixel i-1 is more curved, the tangent and tangent The larger the angle between them, the greater the curvature at pixel i-1, and when the number of corner points around these two points increases, The larger the value of ; when the edge curve between the two points is more curved, the smaller the Euclidean distance between the two points is. The larger the value of The larger the value of , the greater the curvature of the edge curve between pixel i and pixel i-1. The curvature mutation degree of the edge curve of sub-region r is calculated as follows: Where, is the curvature mutation degree of the edge curve of sub-region r, N is the number of pixels on the edge curve of sub-region r, is the curvature evaluation factor between pixel i and the adjacent previous pixel i-1 on the edge curve, and u is the average of the curvature evaluation factors between all pixels except the first pixel and the adjacent previous pixel on the edge curve of subregion r.

[0040] If the curvature of the edge curve of sub-region r at each pixel changes very drastically and frequently, The larger the value of , the more rugged the edge of the sub-region r is, and the greater the possibility that the sub-region r is a fractured area. The smaller the value of , the smoother the edge of the sub-region r is, and the less likely the sub-region r is a fractured area.

[0041] Due to the high-precision manufacturing process and complex and diverse structural features of precision stamping parts for new energy vehicles, some precision stamping parts often exhibit significant variations in edge curvature. However, even for these parts with significant curvature variations, the curvature distribution of edge pixels exhibits regularity and continuity. In contrast, analysis of crack propagation at fracture sites, driven by energy loss patterns and stress concentration, reveals that crack widths are typically greatest closest to the edge of the precision stamping part and gradually decrease with increasing distance from the edge. Furthermore, the crack curvature formed during this process exhibits no repeatability and exhibits irregular variations.

[0042] The curvature evaluation factors between all pixels except the first pixel on the edge curve and the adjacent previous pixel constitute the curvature sequence of the edge curve, which is recorded as At the same time, the Euclidean distances between all pixels except the first pixel on the edge curve and the adjacent previous pixel constitute the distance sequence of the edge curve, which is recorded as ;in, Represents the curvature evaluation factor between pixel point N and the adjacent previous pixel point N-1 on the edge curve; Represents the Euclidean distance between pixel N and the adjacent previous pixel N-1 on the edge curve. Calculate the curvature sequence and distance sequence The absolute value of the Pearson correlation coefficient between ,like The larger the value of , the higher the correlation between the two sequences, and the more likely there is a strong correlation between the curvature change and the Euclidean distance change; therefore, it can also be explained that in the process of the change of the curvature at each pixel point on the edge curve, the Euclidean distance may also show corresponding changes, proving that there is a certain regularity between the curvature and the Euclidean distance, and that there is a certain regularity in the edge shape change of the sub-region.

[0043] Define a sliding window of length w for the curvature sequence The sliding window processing is performed, and a sliding step is recorded as step, to obtain N-20 subsequences with a length of w. The values of w and step can be set by the implementer, and in the embodiment, w is 20 and step is 1.

[0044] The Pearson correlation coefficients between the first subsequence and each of the other subsequences are calculated respectively, and the mean value is recorded as The greater the value of , the higher the correlation between the first subsequence and many subsequences, which means that the curvature change has a certain regularity. According to the Pearson correlation coefficient between the curvature sequence and the distance sequence , the absolute value of the Pearson correlation coefficient and are used to construct the fracture edge fitting degree of the sub-region r, and the expression is as follows: In the expression, the fracture edge fitting degree of the sub-region r is , the Pearson correlation coefficient between the curvature sequence and the distance sequence is , and the mean value of the Pearson correlation coefficients between the first subsequence and all other subsequences is .

[0045] The greater the value of , the stronger the regularity of the edge curvature distribution of the sub-region r, and the greater the value of ; when there is a stronger correlation between the curvature change and the Euclidean distance change between each pixel point on the edge curve and the adjacent previous pixel point, the edge of the sub-region is more likely to be the edge of a good stamping part, and the greater the value of , the greater the value of ; therefore, the greater the value of , the greater the possibility that the sub-region r is a non-fracture region.

[0046] All non-stamping part regions in the image are set as background regions. Since the fracture is caused by the local stress of the plate material being too large, resulting in material failure, around the fracture edge, some stretched texture features can be clearly observed, which are caused by the stretching deformation of the material under stress, and are manifested as elongated textures extending along the fracture edge. Although the stamping part is a stamping part produced by the plate material under the external force of the stamping machine, the entire stamping part has stretched texture features; however, if a fracture defect occurs, the stress on the plate material around the defect will be greater than that in other intact places, and the stretched texture features around the defect will be significantly higher than those in the non-defect part. Since the stamping part has already fractured, when the image is taken, the background region will be captured through the gap of the fracture defect. Due to the obstruction of the stamping part, when the stamping part partially blocks the light, the gray value of the pixel points at the edge of the fracture region of the stamping part will be lower, regardless of the color of the stamping machine bed below the stamping part.

[0047] Still taking the sub-region r as an example, a L2*L2 window is constructed with each pixel point on the edge curve of the sub-region r as the center, and the value of L2 can be set by the implementer, and in this embodiment, L2 is 5. The MLBP value of each pixel point on the edge curve of the sub-region r is obtained by using the MLBP algorithm. The local stress discontinuity index of the sub-region r is calculated, and the expression is as follows: In the formula, is the local stress discontinuity index of the sub-region r, N is the number of pixel points on the edge curve of the sub-region r, is the gray value of the pixel point i on the edge curve of the sub-region r; is the mean value of the gray values of all pixel points on the edge curve of the sub-region r, is the MLBP value of the pixel point i on the edge curve of the sub-region r, is the curvature evaluation factor between the pixel point i and the adjacent previous pixel point i-1 on the edge curve.

[0048] represents the degree of change of the edge gray value of the sub-region r, and reflects the unevenness of the stress distribution; if the sub-region r is a fracture region, it will have very obvious stretching features, and the value is larger; at this time, the greater the difference in the gray values of the pixel points on the edge curve of the sub-region r and the more obvious the stretching features, and the greater the bending degree of the edge curve at each pixel point, the more in line with the characteristics of the fracture region, the value is larger.

[0049] The fracture region confidence of the sub-region r is calculated through the curvature mutation degree of the edge of the sub-region r, the fracture edge fitting degree and the local stress discontinuity index, and the expression is as follows: In the formula, a fracture region confidence of a sub-region r, a fracture edge fitness of a sub-region r, a preset tuning coefficient greater than zero, to avoid a denominator of 0, The value of the tuning coefficient can be set by the implementer, and in the embodiment, the tuning coefficient is 0.001. The tuning coefficient is 0.001.

[0050] The higher the fracture region confidence of the sub-region r, the greater the possibility that the sub-region r is a fracture region. The acquisition process of the fracture region confidence is shown in FIG. 2.

[0051] In step S003, the fracture regions are clustered and divided using the k-means clustering algorithm according to the fracture region confidence of all sub-regions.

[0052] According to the method of calculating the fracture region confidence of the sub-region r, the fracture region confidence of each of the remaining sub-regions in the complete stamping part gray-scale image is calculated. The fracture region confidence of all sub-regions in the complete stamping part gray-scale image is used as input, and the k-means clustering algorithm is used to obtain K clusters, and the value of K can be set by the implementer, and in the embodiment, K is 4. The k-means clustering algorithm is a known technology, and will not be described in detail in the embodiment. The mean value of all fracture region confidences in the four clusters is calculated respectively, and when the normalized value of the mean value of the cluster is greater than the preset fracture region confidence threshold, the sub-region corresponding to the cluster is determined as a fracture region, otherwise, the sub-region corresponding to the cluster is determined as a non-fracture region. The value of the fracture region confidence threshold can be set by the implementer, and in the embodiment, the fracture region confidence threshold is 0.8.

[0053] Thus, the rapid detection of the fracture region of the new energy automobile stamping precision part can be realized.

[0054] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. The above describes a specific embodiment of the present application. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are possible or can be advantageous.

[0055] Each embodiment in the specification is described in a progressive manner, and the same or similar parts of each embodiment can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0056] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than limit them; the technical solutions recorded in the foregoing embodiments are modified, or some technical features are replaced equivalently, and the essence of the corresponding technical solutions does not deviate from the scope of the technical solutions of the embodiments of the present application, which should be included in the protection scope of the present application.

Claims

1. A method for rapid detection of fracture areas of stamped precision parts for new energy vehicles, characterized in that: The method comprises the following steps: Collect images of stamping precision parts and perform grayscale processing on the images to obtain grayscale images of complete stamping parts; The watershed segmentation algorithm is used to obtain each sub-region in the grayscale image of the complete stamping part, and the edge detection algorithm is used to detect the edge of the sub-region. The fitting curve of all edge pixel points of the sub-region is used as the edge curve of the sub-region. The curvature evaluation factor between each pixel point on the edge curve and the adjacent previous pixel point is constructed according to the neighborhood pixel points of each pixel point on the edge curve; the curvature mutation degree of the edge curve is constructed according to the curvature evaluation factor between each pixel point on the edge curve and the adjacent previous pixel point; the curvature sequence of the edge curve is constructed according to the curvature evaluation factor between each pixel point on the edge curve and the adjacent previous pixel point; the distance sequence of the edge curve is constructed according to the coordinates of the pixel points on the edge curve; and the edge curve is obtained. The curvature sequence of each subsequence is constructed; the fracture edge fit of the sub-region is constructed according to the curvature sequence, distance sequence and each subsequence of the curvature sequence of the edge curve; a square window is constructed with each pixel point on the edge curve of the sub-region as the center, and the MLBP algorithm is used to calculate the MLBP value of the square window center; the local stress discontinuity index of the sub-region is constructed according to the grayscale value, MLBP value and curvature evaluation factor of the pixel point on the edge curve of the sub-region; the fracture area confidence of the sub-region is constructed according to the local stress discontinuity index of the sub-region, the fracture edge fit and the curvature mutation degree of the edge curve of the sub-region; according to the fracture area confidence of all sub-regions, a clustering algorithm is used to cluster and divide the fracture area.

2. A rapid detection method for fracture areas of stamped precision parts for new energy vehicles as claimed in claim 1, characterized in that: The step of constructing a curvature evaluation factor between each pixel point on the edge curve and an adjacent previous pixel point based on the neighboring pixel points of each pixel point on the edge curve includes: The least squares method is used to calculate the tangent line at each pixel on the edge curve, and a square window is constructed with each pixel on the edge curve as the center. The curvature evaluation factor between the pixel i on the edge curve and the adjacent previous pixel i-1 is The expression is: Where, is the tangent line at pixel i The angle between In order to tangent and tangent Convert the angle between them into radians. is the Euclidean distance between pixel i and the adjacent previous pixel i-1, is the number of corner points in the square window centered at pixel i, is the number of corner points in the square window centered at pixel i-1.

3. A rapid detection method for fracture areas of stamped precision parts for new energy vehicles as claimed in claim 1, characterized in that: The step of constructing the curvature mutation degree of the edge curve according to the curvature evaluation factor between each pixel point and the adjacent previous pixel point on the edge curve includes: Calculate the average of the curvature evaluation factors between all the pixels except the first pixel on the edge curve and the adjacent previous pixel, calculate the absolute value of the difference between the curvature evaluation factors of all the pixels except the first pixel and the adjacent previous pixel on the edge curve minus the average, and calculate the average of the absolute values ​​of the differences between all the pixels except the first pixel and the adjacent previous pixel on the edge curve as the degree of curvature mutation of the edge curve.

4. A method for rapid detection of fracture areas of stamped precision parts for new energy vehicles as claimed in claim 1, characterized in that: The method of constructing a curvature sequence of an edge curve based on a curvature evaluation factor between each pixel point on the edge curve and an adjacent previous pixel point includes: forming a curvature sequence of the edge curve by combining the curvature evaluation factors between all pixel points except the first pixel point on the edge curve and the adjacent previous pixel point.

5. A rapid detection method for fracture areas of stamped precision parts for new energy vehicles as claimed in claim 1, characterized in that: The step of constructing the distance sequence of the edge curve according to the coordinates of the pixel points on the edge curve includes: forming the distance sequence of the edge curve by combining the Euclidean distances between all the pixel points except the first pixel point and the adjacent previous pixel point on the edge curve.

6. A rapid detection method for fracture areas of stamped precision parts for new energy vehicles as claimed in claim 1, characterized in that: The obtaining of subsequences of the curvature sequence of the edge curve includes: dividing the curvature sequence into subsequences using a sliding window of a preset length, wherein the sliding step length of the sliding window is a preset step length.

7. A rapid detection method for fracture areas of stamped precision parts for new energy vehicles as claimed in claim 1, characterized in that: The step of constructing the sub-region fracture edge conformity according to the curvature sequence, the distance sequence, and each subsequence of the curvature sequence of the edge curve includes: The absolute value of the Pearson correlation coefficient between the curvature sequence and the distance sequence of the edge curve is calculated, the mean value of the Pearson correlation coefficient between the first subsequence of the curvature sequence and all other subsequences is calculated, and the product of the absolute value and the mean value is used as the fracture edge fit of the subregion.

8. A method for rapid detection of fracture areas of stamped precision parts for new energy vehicles as claimed in claim 1, characterized in that: The constructing of the local stress discontinuity index of the sub-region includes: Calculate the mean of the grayscale values ​​of all pixels on the edge curve of the sub-area, calculate the absolute value of the difference between the grayscale value of each pixel except the first pixel on the edge curve of the sub-area and the mean, calculate the MLBP value of each pixel except the first pixel on the edge curve plus the curvature evaluation factor between each pixel and the adjacent previous pixel, calculate the product of the absolute value of the difference of each pixel except the first pixel on the edge curve of the sub-area and the calculation result, and calculate the mean of the products of all pixel points on the edge curve of the sub-area except the first pixel as the local stress discontinuity index of the sub-area.

9. A method for rapid detection of fracture areas of stamped precision parts for new energy vehicles as claimed in claim 1, characterized in that: The constructing of the fracture region confidence of the sub-region includes: Calculate the sum of the curvature mutation degree of the edge curve of the sub-region and the local stress discontinuity index of the sub-region, which is recorded as the first sum; calculate the sum of the fracture edge fit of the sub-region and the preset parameter adjustment coefficient greater than zero, which is recorded as the second sum; and use the ratio of the first sum to the second sum as the fracture area confidence of the sub-region.

10. A rapid detection method for fracture areas of stamped precision parts for new energy vehicles as claimed in claim 1, characterized in that: The method of clustering and dividing the fractured areas using a clustering algorithm based on the fractured area confidence of all sub-areas includes: The fracture area confidence of all sub-areas in the grayscale image of the complete stamping part is used as input, and a clustering algorithm is used to obtain a preset number of clusters. The mean of the fracture area confidence of each cluster is calculated. When the normalized value of the mean of the cluster is greater than the preset fracture area confidence threshold, the sub-area corresponding to the cluster is determined as a fracture area; otherwise, the sub-area corresponding to the cluster is determined as a non-fracture area.

Images