A method, device and system for quickly detecting burrs in cable joints

Abstract

The present invention relates to the technical field of image defect recognition, and in particular to a method, device and system for rapid detection of burrs in cable joints. The present invention calculates the corresponding edge feature abnormality based on the characteristics that the edges of burr areas are usually sharp and the corner points are densely distributed and tend to be in the same direction; further, based on the characteristics that the connected domain feature changes of the burr area at different shooting angles are greater than those of the normal area, the abnormality degree characterizing the abnormal features of the burr area is determined, so that all abnormal burr areas are screened out according to the abnormality degree; finally, based on the similarity between each connected domain and the abnormal burr area in overall grayscale and edge gradient and the abnormal burr area, the possibility of the burr area of ​​each connected domain is further calculated in combination with the edge feature abnormality, so that burr detection is performed based on the burr area possibility and the abnormal burr area that has been screened out, so that the accuracy of the detected real burr area is higher.

Description

Technical Field

[0001] The present invention relates to the technical field of image defect recognition, and in particular to a method, device and system for rapid detection of burrs on cable joints. Background Art

[0002] The burrs at the cable joints will damage the insulation performance of the cable, affect the energy efficiency of the electrical appliances, increase the risk of failure, and interfere with the transmission of signals to a certain extent; therefore, the presence of burrs at the cable joints will directly affect the performance and safety of the cable. Therefore, it is necessary to accurately and timely detect the burrs at the cable joints, so as to eliminate the burrs in time to improve the performance and safety of the cable. Considering that the efficiency and accuracy of manual detection of cable joint burrs are low, image processing methods are usually used to automatically detect burrs to ensure the efficiency and accuracy of cable joint burrs.

[0003] The prior art usually uses a deep learning-based target detection method to identify the burr area in the cable cross-section image at the cable joint. Since the deep learning-based target detection method needs to use a deep learning model to identify the burr area, a large number of training samples are required to ensure the accuracy of the recognition effect; in addition, the burr area has a certain randomness in distribution and diverse shapes, and the deep learning-based target detection method may not be able to fully capture the burr features, resulting in some burr areas being missed. Therefore, the prior art uses a deep learning-based target detection method to detect burr areas with low accuracy. Summary of the invention

[0004] The present application provides a method, device and system for rapid detection of burrs in cable joints. According to the characteristics that the edges of the burr area are usually sharp, the density of corner points is large, and the directional distribution of the corner points is close to the same, the similarity of each connected domain with the burr area in terms of morphological edge is characterized by the calculated edge feature abnormality; further, based on the edge feature abnormality, according to the characteristic that the connected domain feature changes of the burr area at different shooting angles are greater than those of the normal area, the abnormality degree of the abnormal feature characterizing the abnormal feature of the burr area corresponding to each connected domain under the main viewing angle is determined, so that all abnormal burr areas are screened out according to the abnormality degree; finally, according to the similarity between each connected domain and the abnormal burr area in terms of overall grayscale and edge gradient, combined with the edge feature abnormality, the possibility of the burr area of ​​each connected domain is further calculated, so that burr detection is performed based on the possibility of the burr area and the abnormal burr area that has been screened out, so as to solve the problem that the target detection method used in the prior art may not be able to fully capture the burr features, so that the accuracy of the detected real burr area is higher.

[0005] The first aspect of the present application provides a method for quickly detecting burrs in a cable joint, comprising:

[0006] Capturing the cable cross-section grayscale image of the cable cross-section at each shooting angle at the cable joint by a camera; acquiring all connected domains in the cable cross-section grayscale image; the shooting angles include a main viewing angle and at least two auxiliary viewing angles;

[0007] In the cable cross-section grayscale image, the edge feature abnormality of each connected domain is determined according to the similarity between the shape of each connected domain and a circle and the consistency between the corner point density and the direction distribution of the corner points in the connected domain;

[0008] According to the similarity in overall gray value, edge feature abnormality and position between each connected domain in the main viewing angle and all connected domains in each auxiliary viewing angle, the matching degree between each connected domain in the main viewing angle and each connected domain in each auxiliary viewing angle is determined;

[0009] Determine the matching connected domains of each connected domain in the main viewing angle in each auxiliary viewing angle according to the matching degree; determine the abnormality degree of each connected domain in the main viewing angle according to the abnormality degree of edge features of each connected domain in the main viewing angle and the fluctuation of the matching degree of each corresponding matching connected domain; and screen out all abnormal burr areas in the main viewing angle according to the abnormality degree;

[0010] According to the similarity between each connected domain and each abnormal burr area in grayscale gradient and grayscale and the corresponding edge feature abnormality under the main viewing angle, the possibility of the burr area of ​​each connected domain under the main viewing angle is determined; according to the burr area possibility and all abnormal burr areas, all real burr areas are detected.

[0011] In an embodiment of the first aspect of the present application, the method for obtaining the edge feature abnormality includes:

[0012] Calculating the circularity of each connected domain in the cable cross-section grayscale image;

[0013] Determine all corner points corresponding to each connected domain and the corner point direction of each corner point by using the corner point detection method; quantify the corner point direction of each corner point into a corner point direction angle according to the angle between the corner point direction of each corner point and the horizontal direction; calculate the corner point direction angle variance corresponding to all corner points in each connected domain;

[0014] Determine the corner point density of each connected domain according to the number of corner points in each connected domain and the area of ​​the connected domain;

[0015] The edge feature abnormality of each connected domain is determined according to the corner point direction angular variance, the corner point density and the circularity; the corner point direction angular variance and the circularity are both negatively correlated with the edge feature abnormality, and the corner point density is positively correlated with the edge feature abnormality.

[0016] In an embodiment of the first aspect of the present application, the method for obtaining the matching degree includes:

[0017] Determine the pixel coordinates of the centroid point of each connected domain in the cable cross-section grayscale image and calculate the grayscale value mean of all pixels in each connected domain; take each connected domain under the main viewing angle as the target connected domain in turn; take each connected domain under each auxiliary viewing angle as the auxiliary connected domain in turn;

[0018] Determining a corresponding reference grayscale difference according to a difference between a grayscale value mean of the target connected domain and a grayscale value mean of the auxiliary connected domain;

[0019] Determining a corresponding reference position difference according to a Euclidean distance between pixel coordinates of a centroid point of the target connected domain and pixel coordinates of a centroid point of the auxiliary connected domain;

[0020] Determining a corresponding edge anomaly difference according to a difference between an edge feature anomaly degree of the target connected domain and an edge feature anomaly degree of the auxiliary connected domain;

[0021] The matching degree between the target connected domain and the auxiliary connected domain is determined according to the reference grayscale difference, the reference position difference and the edge anomaly difference; the reference grayscale difference, the reference position difference and the edge anomaly difference are all negatively correlated with the matching degree.

[0022] In an embodiment of the first aspect of the present application, the method for obtaining the matching connected domain includes:

[0023] The connected domain corresponding to the maximum matching degree between each auxiliary perspective and the target connected domain is used as the matching connected domain of the target connected domain in each auxiliary perspective.

[0024] In an embodiment of the first aspect of the present application, the method for obtaining the abnormality degree includes:

[0025] According to the matching degree variance between the target connected domain and all corresponding matching connected domains, the matching degree discreteness of the target connected domain is determined; according to the matching degree discreteness and the edge feature abnormality corresponding to the target connected domain, the abnormality degree of the target connected domain is determined; the matching degree discreteness and the edge feature abnormality corresponding to the target connected domain are both positively correlated with the abnormality degree.

[0026] In an embodiment of the first aspect of the present application, the method for acquiring the abnormal burr area includes:

[0027] Among all connected domains under the main viewing angle, the connected domains with abnormality greater than a preset abnormality threshold are regarded as abnormal burr regions.

[0028] In an embodiment of the first aspect of the present application, the method for obtaining the possibility of the burr area includes:

[0029] Determine the reference abnormal gray value according to the average gray value of all pixels in all abnormal burr areas; determine the reference abnormal gradient value according to the average gray gradient value of all edge pixels in all abnormal burr areas;

[0030] The difference between the mean grayscale value of all pixels in each connected domain under the main viewing angle and the reference abnormal grayscale value is subjected to negative correlation mapping processing to determine the grayscale abnormality degree of each connected domain under the main viewing angle;

[0031] The difference between the mean grayscale gradient value of all edge pixels in each connected domain under the main viewing angle and the reference abnormal gradient value is subjected to negative correlation mapping processing to determine the degree of gradient abnormality of each connected domain under the main viewing angle;

[0032] According to the grayscale abnormality, the gradient abnormality and the edge feature abnormality corresponding to each connected domain under the main viewing angle, the possibility of the burr area of ​​each connected domain under the main viewing angle is determined; the edge feature abnormality corresponding to each connected domain under the main viewing angle, the grayscale abnormality and the gradient abnormality are all positively correlated with the possibility of the burr area.

[0033] In an embodiment of the first aspect of the present application, the method for obtaining the real burr area includes:

[0034] Among all connected domains under the main viewing angle, connected domains whose burr region probability is greater than a preset burr threshold and all abnormal burr regions are regarded as real burr regions.

[0035] In a second aspect, the present application provides a cable joint burr rapid detection system, the system comprising:

[0036] A data acquisition preprocessing module is used to acquire a grayscale image of a cable cross section at each shooting angle through a camera; and obtain all connected domains in the grayscale image of the cable cross section; the shooting angle includes a main viewing angle and at least two auxiliary viewing angles;

[0037] A first determination module is used to determine the abnormality of the edge features of each connected domain in the cable cross-section grayscale image according to the similarity between the shape of each connected domain and a circle and the consistency of the corner point density and the direction distribution of the corner points in the connected domain;

[0038] The second determination module is used to determine the matching degree between each connected domain under the main viewing angle and each connected domain under each auxiliary viewing angle according to the similarity in overall gray value, edge feature abnormality and position between each connected domain under the main viewing angle and all connected domains under each auxiliary viewing angle;

[0039] The abnormal burr area screening module is used to determine the matching connected domains of each connected domain in the main viewing angle in each auxiliary viewing angle according to the matching degree; determine the abnormal degree of each connected domain in the main viewing angle according to the abnormal degree of edge features of each connected domain in the main viewing angle and the fluctuation of the matching degree of each corresponding matching connected domain; and screen out all abnormal burr areas in the main viewing angle according to the abnormal degree;

[0040] The real burr area detection module is used to determine the burr area possibility of each connected domain under the main viewing angle according to the similarity between each connected domain under the main viewing angle and each abnormal burr area in grayscale gradient and grayscale and the corresponding edge feature abnormality; all real burr areas are detected according to the burr area possibility and all abnormal burr areas.

[0041] In a third aspect, the present application provides a cable joint burr rapid detection device, comprising a memory and a processor. The memory is used to store computer program code, and the processor is used to call and run the computer program code from the memory to execute the method of the first aspect of the present application or any embodiment of the first aspect.

[0042] In a fourth aspect, the present application provides a computer program product, comprising a computer program code, which, when executed, performs the method of the first aspect of the present application or any embodiment of the first aspect.

[0043] In a fifth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program code. When the computer program code is executed, the method of the first aspect or any embodiment of the first aspect of the present application is performed.

[0044] This application has the following beneficial effects:

[0045] In the dimension of the edge shape of the burr area, the edge of the burr area is sharper than that of other normal areas, that is, it is more different from a circle. The corresponding corner point density is greater than that of the normal area and the directions corresponding to these corner points show a higher consistency. Therefore, the present application first calculates the edge feature abnormality corresponding to each edge domain in the dimension of the morphological edge based on the similarity of the shape of each connected domain to a circle and the consistency of the corner point density and the direction distribution of the corner points in the connected domain.

[0046] Furthermore, since the edges of the burr area are more prominent and the physical structure of the burr is usually affected by light to a certain extent, the feature performance of the connected domain corresponding to the burr on the image under different shooting angles varies more than that of the connected domain of the normal cable area, i.e., the non-burr area. Therefore, it is necessary to calculate the degree of abnormality based on the feature changes of each connected domain in each auxiliary perspective under the main perspective combined with the edge feature abnormality, so as to calculate a more accurate abnormal burr area based on the degree of abnormality.

[0047] Finally, in order to avoid the problem that some burr areas are not screened out due to shooting angle problems, the possibility of the burr area of ​​each connected domain is further calculated based on the similarity between each connected domain and the abnormal burr area in overall grayscale and edge gradient combined with the edge feature abnormality. Burr detection is then performed based on the possibility of the burr area and the abnormal burr area that has been screened out, so that the accuracy of the detected real burr area is higher. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0049] Figure 1 A flow chart of a method for rapid detection of burrs in cable joints provided by one embodiment of the present invention;

[0050] Figure 2 A schematic diagram of camera position and shooting angle provided by an embodiment of the present invention;

[0051] Figure 3 A flow chart of a method for obtaining an edge feature abnormality degree of a cable joint burr rapid detection method provided by an embodiment of the present invention;

[0052] Figure 4 A flow chart of a method for obtaining a matching degree of a cable joint burr rapid detection method provided by an embodiment of the present invention;

[0053] Figure 5 A flow chart of a method for obtaining a burr area possibility of a cable joint burr rapid detection method provided by an embodiment of the present invention;

[0054] Figure 6 A structural diagram of a cable joint burr rapid detection system provided by one embodiment of the present invention;

[0055] Figure 7 A schematic structural diagram of a device for rapid detection of burrs in cable joints provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0056] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a cable joint burr rapid detection method, device and system proposed in accordance with the present invention, its specific implementation method, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment, and the specific features, structures or characteristics in one or more embodiments may be combined in any suitable form. In addition, the terms "first" and "second" are used only for descriptive purposes and cannot be understood as implying or suggesting relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features.

[0057] Unless defined otherwise, 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 invention belongs.

[0058] The specific scheme of a cable joint burr rapid detection method, device and system provided by the present invention is described in detail below with reference to the accompanying drawings.

[0059] The present application provides a method for quickly detecting burrs in cable connectors. Figure 1 , which shows a flow chart of a method for rapid detection of cable joint burrs provided by an embodiment of the present invention, the method comprising:

[0060] Step S101: using a camera to capture a cable cross-section grayscale image of the cable cross-section at each shooting angle; obtaining all connected domains in the cable cross-section grayscale image; the shooting angle includes a main viewing angle and at least two auxiliary viewing angles.

[0061] Cable burrs are usually caused by improper operation when cutting or stripping the cable. These burrs usually appear at the cross section of the cable connector. Therefore, it is necessary to collect the image of the cable cross section, that is, the approximately circular cut section of the cable. In a specific implementation of the embodiment of the present invention, the cable cross section at the cable connector is placed vertically upward, and three high-precision industrial cameras are set at different positions at the same height above the cable cross section, so that each high-precision industrial camera shoots the cable cross section image at each shooting angle toward the cable cross section, and the cable cross section image is grayed to obtain the cable cross section grayscale image required by the embodiment of the present invention; please refer to Figure 2, which shows a schematic diagram of camera position and shooting angle provided by an embodiment of the present invention; Figure 2 The cable cross section being recorded is present in And three high-precision industrial cameras correspond to , and ; and a high-precision industrial camera directly above the cable cross section The corresponding shooting angle is the main viewing angle, and the other two high-precision industrial cameras and Corresponding to the auxiliary perspective; and in order to improve the rigor of subsequent analysis, high-precision industrial cameras and and The distances between them are equal and they are located on the same straight line.

[0062] Further considering that the burr area is an independent area, its internal grayscale value is different from that of the surrounding area, so the burr area is usually a connected domain, so first obtain all the connected domains in the cable cross-section grayscale image. In a specific implementation of the embodiment of the present invention, the cable cross-section grayscale image is segmented into superpixels, and each superpixel block obtained is used as the connected domain required by the embodiment of the present invention. In other specific implementations of the embodiment of the present invention, the cable cross-section grayscale image can also be edge detected by the canny edge detection method, and each closed area surrounded by the edge is used as a connected domain; all connected domains can also be determined by a regional growing algorithm. It should be noted that superpixel segmentation, canny edge detection method and regional growing algorithm are all commonly used technical means well known to those skilled in the art, and implementers can also obtain connected domains through other connected domain acquisition methods, which will not be further limited or elaborated here.

[0063] Step S102: in the cable cross-section grayscale image, the edge feature abnormality of each connected domain is determined according to the similarity between the shape of each connected domain and a circle and the consistency between the corner point density and the direction distribution of the corner points in the connected domain.

[0064] In the dimension of the edge shape of the burr area, the edge of the burr area is sharper than that of other normal areas, that is, it is more different from a circle. The corresponding corner point density is greater than that of the normal area and the directions corresponding to these corner points show a higher consistency. Therefore, this application first calculates the edge feature abnormality corresponding to each edge domain in the dimension of the morphological edge based on the similarity of the shape of each connected domain to a circle and the consistency of the corner point density and the direction distribution of the corner points in the connected domain.

[0065] Step S103: Determine the degree of matching between each connected domain in the main viewing angle and each connected domain in each auxiliary viewing angle according to the similarity in overall grayscale value, edge feature abnormality and position between each connected domain in the main viewing angle and all connected domains in each auxiliary viewing angle.

[0066] Step S104: Determine the matching connected domains of each connected domain in the main perspective in each auxiliary perspective according to the matching degree; determine the abnormality degree of each connected domain in the main perspective according to the abnormality degree of edge features of each connected domain in the main perspective and the fluctuation of the matching degree of each corresponding matching connected domain; and screen out all abnormal burr areas in the main perspective according to the abnormality degree.

[0067] Furthermore, since the edges of the burr area are more prominent and the physical structure of the burr is usually affected by light to a certain extent, the feature performance of the connected domain corresponding to the burr on the image under different shooting angles varies more than that of the connected domain of the normal cable area, i.e., the non-burr area. Therefore, it is necessary to calculate the degree of abnormality based on the feature changes of each connected domain in each auxiliary perspective under the main perspective combined with the edge feature abnormality, so as to calculate a more accurate abnormal burr area based on the degree of abnormality.

[0068] Since it is necessary to analyze the changes of each connected domain under different shooting angles, it is necessary to match the connected domains under different shooting angles first, that is, to clarify the matching connected domains corresponding to each connected domain in the main view in each auxiliary view. For each connected domain under the main view, although the feature performance of the corresponding matching connected domain under the auxiliary view will change to a certain extent, the matching connected domain should be more similar in grayscale, position and edge feature abnormality than other connected domains under the auxiliary view. Therefore, firstly, the matching degree between each connected domain under the main view and each connected domain under each auxiliary view is determined based on the similarity in overall grayscale value, edge feature abnormality and position, and then the matching connected domain of each connected domain under the main view in each auxiliary view is determined based on the matching degree. Further, according to the characteristics that the feature performance of the connected domain corresponding to the burr under different shooting angles changes more greatly on the image, and the edge feature abnormality of the burr area is usually larger, the corresponding abnormality is determined based on the fluctuation between the matching degrees of each matching connected domain corresponding to each connected domain under the main view and the edge feature abnormality of each connected domain. The greater the difference observed between each auxiliary perspective for each connected domain under the main perspective, and the greater the abnormality of the edge features of each connected domain under the main perspective, the greater the corresponding degree of abnormality, that is, the more likely it is to belong to the burr abnormal area, and a more accurate burr abnormal area can be screened out according to the degree of abnormality.

[0069] Step S105: Determine the burr region possibility of each connected domain under the main viewing angle according to the similarity between each connected domain under the main viewing angle and each abnormal burr region in grayscale gradient and grayscale and the corresponding edge feature abnormality; detect all real burr regions according to the burr region possibility and all abnormal burr regions.

[0070] Finally, in order to avoid the problem that some burr areas are not screened out due to shooting angle problems, the possibility of the burr area of ​​each connected domain is further calculated based on the similarity between each connected domain and the abnormal burr area in overall grayscale and edge gradient combined with the edge feature abnormality. Burr detection is then performed based on the possibility of the burr area and the abnormal burr area that has been screened out, so that the accuracy of the detected real burr area is higher.

[0071] Preferably, in some possible implementations of the embodiments of the present invention, the specific implementation process of the method for obtaining the edge feature abnormality in step S102 includes: Figure 3 , which shows a flow chart of a method for obtaining edge feature abnormality of a cable joint burr rapid detection method provided by an embodiment of the present invention, the method comprising:

[0072] Step S301: Calculate the circularity of each connected domain in the cable cross-section grayscale image.

[0073] According to the objective morphological manifestation of the burr region, the corresponding burr region usually exhibits a sharp or protruding edge morphology. Therefore, the larger the corresponding circularity, the more likely the corresponding connected domain does not belong to the burr region.

[0074] In a specific implementation of the embodiment of the present invention, the process of obtaining the circularity is expressed by the formula: ;in, Connected domain Circularity; Connected domain area; Connected domain circumference; It should be noted that circularity is a technical means well known to those skilled in the art, and its significance will not be further elaborated here.

[0075] In another specific implementation of the embodiment of the present invention, the method for obtaining the circularity includes: calculating the connected domain Each edge pixel to the connected domain The Euclidean distance between the centroids of the connected domain is the normalized value of the reciprocal of the sum of the variances of all Euclidean distances and the positive number 1. The circularity of all edge pixels to the connected domain The closer the Euclidean distance between the centroids is, that is, the smaller the corresponding variance is, the closer the connected domain is to a circle; therefore, the corresponding circularity is determined by normalizing after negative correlation mapping through the inverse.

[0076] Step S302: determine all corner points corresponding to each connected domain and the corner point direction of each corner point by means of a corner point detection method; quantify the corner point direction of each corner point into a corner point direction angle according to the angle between the corner point direction of each corner point and the horizontal direction; calculate the corner point direction angle variance corresponding to all corner points in each connected domain.

[0077] Since the edge of the burr area is relatively sharp, that is, its edge is usually not smooth, there are usually more corner points on the edge of the burr area; and since the burr area usually extends in a certain direction, the corresponding corner point direction will show a high consistency; therefore, the smaller the variance of the corner point direction angles of all corner points corresponding to the connected domain, the more consistent the corner point directions are, and the higher the possibility that the connected domain belongs to the burr area. It should be noted that the corner point direction in the embodiment of the present invention is the gradient direction of the pixel point corresponding to the corner point. It should be noted that the implementer can replace the horizontal direction with any direction according to the specific implementation environment. The horizontal direction is only used as a reference to quantify the corner point direction, and will not be further elaborated here.

[0078] Step S303: Determine the corner point density of each connected domain according to the number of corner points in each connected domain and the area of ​​the connected domain.

[0079] As a defect, the burr area usually has a smaller area. Combined with the fact that it has a large number of corner points, the greater the corner point density of the connected domain, the greater the possibility that the corresponding connected domain belongs to the burr area.

[0080] Step S304: Determine the edge feature abnormality of each connected domain according to the corner point direction angle variance, corner point density and circularity.

[0081] Since the smaller the angle variance of the corner point direction, the smaller the circularity, and the greater the corner point density, the greater the possibility that the corresponding connected domain belongs to the burr area in the dimension of the edge shape. Therefore, the angle variance and circularity of the corner point direction are negatively correlated with the abnormality of the edge feature, and the corner point density is positively correlated with the abnormality of the edge feature.

[0082] In a specific implementation of the embodiment of the present invention, the process of obtaining the edge feature abnormality is expressed by the formula: ;in, Connected domain The abnormality of edge features; Connected domain Circularity; Connected domain The number of corner points in ; Connected domain The area of ​​the connected domain; Connected domain The angle variance of the corner point directions corresponding to all corner points in ; is an exponential function with a natural constant as the base. Implementers can adopt other negative correlation mapping methods according to the specific implementation environment, such as , and the reciprocal, where is a hyperbolic tangent function, which will not be further described here.

[0083] Preferably, in some possible implementations of the embodiments of the present invention, the specific implementation process of the method for obtaining the matching degree in step S103 includes: Figure 4 , which shows a flow chart of a method for obtaining a matching degree of a cable joint burr rapid detection method provided by an embodiment of the present invention, the method comprising:

[0084] Determine the pixel coordinates of the centroid point of each connected domain in the cable cross-section grayscale image and calculate the grayscale value mean of all pixels in each connected domain; take each connected domain under the main viewing angle as the target connected domain in turn; take each connected domain under each auxiliary viewing angle as the auxiliary connected domain in turn.

[0085] Step S401: Determine a corresponding reference grayscale difference according to the difference between the grayscale value mean of the target connected domain and the grayscale value mean of the auxiliary connected domain.

[0086] For each connected domain under the auxiliary perspective, the more similar its corresponding grayscale feature is to the grayscale feature of the target connected domain, the more likely it is that they belong to the same area after mapping; therefore, the smaller the reference grayscale difference, the greater the corresponding matching degree.

[0087] Step S402: Determine the corresponding reference position difference according to the Euclidean distance between the pixel coordinates of the centroid of the target connected region and the pixel coordinates of the centroid of the auxiliary connected region.

[0088] Under different shooting angles, the relative position of the same connected domain in the overall image does not change, so the corresponding pixel coordinates usually do not change significantly. Therefore, according to the construction principle of pixel coordinates, the closer the pixel coordinates of the centroid points of the target connected domain and the auxiliary connected domain are, the greater the corresponding matching degree; that is, the reference position difference is negatively correlated with the matching degree.

[0089] Step S403: Determine the corresponding edge anomaly difference according to the difference between the edge feature anomaly degree of the target connected component and the edge feature anomaly degree of the auxiliary connected component.

[0090] For each connected domain, although the edge shape of the same connected domain may change to a certain extent under different shooting angles, each connected domain is obviously more similar to its corresponding matching connected domain than other connected domains. Therefore, the closer the corresponding edge feature abnormalities are, that is, the smaller the edge abnormality difference is, the more likely it is that it belongs to the matching connected domain corresponding to each connected domain, and the corresponding matching degree is greater.

[0091] Step S404: determining the matching degree between the target connected domain and the auxiliary connected domain according to the reference grayscale difference, the reference position difference and the edge anomaly difference.

[0092] The matching degree is further characterized by combining the reference grayscale difference, reference position difference and edge anomaly difference. According to the correlation that the larger the reference grayscale difference, the larger the reference position difference and the larger the edge anomaly difference, the smaller the corresponding matching degree, the reference grayscale difference, the reference position difference and the edge anomaly difference are all negatively correlated with the matching degree.

[0093] In a specific implementation of the embodiment of the present invention, the process of obtaining the matching degree is expressed by the formula: ;in, From the main perspective connected domains, For the The first auxiliary perspective connected domains; From the main perspective The connected domain and The first auxiliary perspective The degree of matching between connected domains; From the main perspective The mean gray value of all pixels in a connected domain; For the The first auxiliary perspective The mean gray value of all pixels in a connected domain; From the main perspective The connected domain and The first auxiliary perspective The reference grayscale difference between connected domains; From the main perspective The pixel coordinates of the centroid of the first connected domain are The first auxiliary perspective The Euclidean distance between the pixel coordinates of the centroid points of the connected domains, that is, the corresponding reference position difference; From the main perspective The abnormality of the edge features of the connected domain; For the The first auxiliary perspective The abnormality of the edge features of the connected domain; From the main perspective The connected domain and The first auxiliary perspective The difference in reference positions between connected domains; is the absolute value symbol; is the square root function; It is a linear normalization function. The implementer can adjust the normalization method according to the specific implementation environment.

[0094] In other specific implementations of the embodiments of the present invention, the product of the reference grayscale difference, the reference position difference and the edge anomaly difference may be used to replace The calculation of the matching degree is performed and will not be further elaborated here.

[0095] Preferably, in some possible implementations of the embodiments of the present invention, the specific implementation process of the method for obtaining the matching connected domain in step S104 includes: taking the connected domain corresponding to the maximum matching degree between the target connected domain in each auxiliary perspective as the matching connected domain of the target connected domain in each auxiliary perspective. The greater the matching degree, the more likely it is to belong to the same connected domain, so the connected domain with the maximum matching degree corresponding to the target connected domain is selected as the corresponding matching connected domain; in each auxiliary perspective, each connected domain has a corresponding matching degree with the target connected domain corresponding to the main perspective, so a matching connected domain corresponding to the maximum matching degree can be found in each auxiliary perspective to match the target connected domain; so that the target connected domain in the main perspective obtains a corresponding matching connected domain in each auxiliary perspective.

[0096] Preferably, in some possible implementations of the embodiments of the present invention, the specific implementation process of the method for obtaining the abnormality degree in step S104 includes:

[0097] The matching degree dispersion of the target connected domain is determined according to the matching degree variance between the target connected domain and all corresponding matching connected domains.

[0098] According to the shape characteristics of the burr area, the changes are relatively large compared with other normal connected domains under different viewing angles, resulting in certain differences in the matching degrees of the corresponding matching connected domains under different auxiliary viewing angles, that is, there are certain fluctuation characteristics; while the normal connected domains outside the burr area are usually more rounded, and the corresponding matching degrees under different auxiliary viewing angles have certain similarities; therefore, the discrete degree of matching is determined by the matching degree variance between the target connected domain and all corresponding matching connected domains, so that the larger the discrete degree of matching, the more likely the target connected domain belongs to the burr area. In other possible implementation methods of the embodiments of the present invention, when there are only two auxiliary viewing angles, the square of the difference in matching degree between the target connected domain and the corresponding two matching connected domains is used as the corresponding discrete degree of matching.

[0099] According to the matching degree discreteness and the edge feature abnormality corresponding to the target connected domain, the abnormality of the target connected domain is determined. The greater the abnormality of the edge feature corresponding to the target connected domain under the main viewing angle, the greater the possibility of belonging to the burr area. Therefore, the matching degree discreteness and the edge feature abnormality are combined for comprehensive representation, so that the greater the abnormality obtained, the greater the possibility of the target connected domain under the main viewing angle belonging to the burr area; that is, the matching degree discreteness and the edge feature abnormality corresponding to the target connected domain are positively correlated with the abnormality.

[0100] In a specific implementation of the embodiment of the present invention, the process of obtaining the abnormality degree is expressed by the formula: ;in, From the main perspective The degree of abnormality of the connected domain; From the main perspective The abnormality of the edge features of the connected domain; From the main perspective The variance of the matching degree between the connected domain and all the corresponding matching connected domains, that is, the first The matching degree discreteness of the connected components. It is a linear normalization function. The implementer can adjust the normalization method according to the specific implementation environment. The purpose is to limit the value of the abnormality degree to the range of 0 to 1, which is convenient for screening abnormal burr areas.

[0101] Preferably, in some possible implementations of the embodiments of the present invention, the specific implementation process of the method for obtaining the abnormal burr area in step S104 includes:

[0102] Among all connected domains under the main viewing angle, the connected domains with an abnormality greater than the preset abnormality threshold are regarded as abnormal burr areas. Since the greater the abnormality, the greater the possibility that the corresponding connected domain belongs to the burr area, the preset abnormality threshold is set for screening. In a specific implementation of the embodiment of the present invention, the preset abnormality threshold is set to 0.73, and the implementer can adjust it according to the specific implementation environment.

[0103] Preferably, in some possible implementations of the embodiments of the present invention, the specific implementation process of the method for obtaining the possibility of the burr area in step S105 includes: Figure 5 , which shows a flow chart of a method for obtaining a burr area possibility of a cable joint burr rapid detection method provided by an embodiment of the present invention, the method comprising:

[0104] Step S501: determining a reference abnormal gray value according to the average gray value of all pixels in all abnormal burr regions; determining a reference abnormal gradient value according to the average gray gradient value of all edge pixels in all abnormal burr regions.

[0105] The process of obtaining the possibility of the burr area needs to combine the similarities between each connected domain and the abnormal burr area in terms of overall grayscale and edge gradient; therefore, the reference abnormal grayscale value of the mean grayscale value of all pixels in all abnormal burr areas is calculated as the grayscale standard of the burr area, and the reference abnormal gradient value corresponding to the mean grayscale gradient value is used as the grayscale gradient standard for similarity analysis.

[0106] Step S502: negative correlation mapping is performed on the difference between the grayscale value mean of all pixels in each connected domain under the main viewing angle and the reference abnormal grayscale value to determine the grayscale abnormality degree of each connected domain under the main viewing angle.

[0107] For each connected domain under the main viewing angle, the smaller the difference between its corresponding overall grayscale value, that is, the grayscale value mean, and the reference abnormal grayscale value representing the grayscale standard of the burr area, the more similar the connected domain is to the burr area in the grayscale value dimension; therefore, the greater the degree of grayscale abnormality, the greater the possibility that the corresponding connected domain is a burr area.

[0108] Step S503: negatively correlate the difference between the mean grayscale gradient value of all edge pixels in each connected domain under the main viewing angle and the reference abnormal gradient value to determine the gradient abnormality degree of each connected domain under the main viewing angle.

[0109] For each connected domain under the main viewing angle, the smaller the difference between the overall grayscale gradient value of all corresponding edge pixels, that is, the mean grayscale gradient value and the reference abnormal gradient value representing the grayscale gradient standard of the burr area, the more similar the connected domain is to the burr area in terms of grayscale value dimension; therefore, the greater the degree of gradient abnormality, the greater the possibility that the corresponding connected domain is a burr area.

[0110] Step S504: Determine the possibility of a burr region for each connected domain under the main viewing angle according to the grayscale abnormality, the gradient abnormality and the edge feature abnormality corresponding to each connected domain under the main viewing angle.

[0111] The greater the edge feature abnormality, the more likely the corresponding connected domain is to belong to the burr region in the dimension of the regional edge shape; therefore, the edge feature abnormality is also positively correlated with the possibility of the burr region; that is, the edge feature abnormality, grayscale abnormality and gradient abnormality corresponding to each connected domain under the main viewing angle are all positively correlated with the possibility of the burr region.

[0112] In a specific implementation of the embodiment of the present invention, the possibility of the corresponding burr area is determined according to the product of the edge feature abnormality, grayscale abnormality and gradient abnormality corresponding to each connected domain under the main viewing angle. It should be noted that in addition to the product, the implementer can also calculate the possibility of the burr area by other methods according to the correlation, such as the normalized value of the mean or the normalized value of the sum of the edge feature abnormality, grayscale abnormality and gradient abnormality, etc., which will not be further elaborated here.

[0113] In a specific implementation of the embodiment of the present invention, the process of obtaining the possibility of the burr area is expressed by the formula: ;in, From the main perspective The possibility of burr area in a connected domain; From the main perspective The abnormality of the edge features of the connected domain; From the main perspective The mean gray value of all pixels in a connected domain; is the mean gray value of all pixels in all abnormal burr areas, that is, the reference abnormal gray value; From the main perspective The mean gray gradient value of all edge pixels in a connected domain; is the mean grayscale gradient value of all edge pixels in all abnormal burr areas, that is, the reference abnormal gradient value; From the main perspective The grayscale abnormality of the connected domain; From the main perspective The degree of gradient anomaly of the connected domain; is the absolute value symbol; It is a linear normalization function. The implementer can adjust the normalization method according to the specific implementation environment.

[0114] Preferably, in some possible implementations of the embodiments of the present invention, the specific implementation process of the method for obtaining the real burr area in step S105 includes:

[0115] Among all connected domains under the main viewing angle, the connected domains whose burr region possibility is greater than the preset burr threshold and all abnormal burr regions are taken as real burr regions. Since the greater the possibility of the burr region, the more similar the corresponding connected domain is to the abnormal burr region, and the greater the possibility of belonging to the burr region, the preset burr threshold is set for screening. In a specific implementation of an embodiment of the present invention, the preset burr threshold is set to 0.81, and the implementer can adjust it according to the specific implementation environment. That is, the abnormal burr region is omitted and supplemented by the connected domain whose burr region possibility is greater than the preset burr threshold, so that the obtained real burr region is more accurate.

[0116] In summary, a method for rapid detection of burrs in cable joints calculates the corresponding edge feature abnormality through the characteristics that the edges of the burr area are usually sharp and the corner points are densely distributed and tend to be in the same direction; further, according to the characteristic that the connected domain feature changes of the burr area at different shooting angles are greater than those of the normal area, the abnormality degree characterizing the abnormal features of the burr area is determined, so that all abnormal burr areas are screened out according to the abnormality degree; finally, according to the similarity between each connected domain and the abnormal burr area in overall grayscale and edge gradient, combined with the edge feature abnormality, the possibility of the burr area of ​​each connected domain is further calculated, so that burr detection is performed according to the possibility of the burr area and the abnormal burr area that has been screened out, so that the accuracy of the detected real burr area is higher.

[0117] This application also provides a cable joint burr rapid detection system, please refer to Figure 6 , which shows a structural diagram of a cable joint burr rapid detection system provided by an embodiment of the present invention, the system comprising: a data acquisition preprocessing module 601, a first determination module 602, a second determination module 603, an abnormal burr area screening module 604 and a real burr area detection module 605.

[0118] The data acquisition preprocessing module 601 is used to acquire a grayscale image of the cable cross section at each shooting angle through a camera; obtain all connected domains in the grayscale image of the cable cross section; the shooting angle includes a main viewing angle and at least two auxiliary viewing angles;

[0119] The first determination module 602 is used to determine the abnormality of the edge features of each connected domain in the cable cross-section grayscale image according to the similarity between the shape of each connected domain and the circle and the consistency of the corner point density and the direction distribution of the corner points in the connected domain;

[0120] The second determination module 603 is used to determine the matching degree between each connected domain under the main viewing angle and each connected domain under each auxiliary viewing angle according to the similarity in overall gray value, edge feature abnormality and position between each connected domain under the main viewing angle and all connected domains under each auxiliary viewing angle;

[0121] The abnormal burr region screening module 604 is used to determine the matching connected domains of each connected domain in the main view in each auxiliary view according to the matching degree; determine the abnormal degree of each connected domain in the main view according to the abnormal degree of edge features of each connected domain in the main view and the fluctuation of the matching degree of each corresponding matching connected domain; and screen out all abnormal burr regions in the main view according to the abnormal degree;

[0122] The real burr area detection module 605 is used to determine the burr area possibility of each connected domain under the main perspective based on the similarity between each connected domain under the main perspective and each abnormal burr area in grayscale gradient and grayscale and the corresponding edge feature abnormality; all real burr areas are detected based on the burr area possibility and all abnormal burr areas.

[0123] It should be noted that the system provided in the above embodiment is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the computer device is divided into different functional modules to complete all or part of the functions described above. In addition, the cable joint burr rapid detection system and the cable joint burr rapid detection method embodiment provided in the above embodiment belong to the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0124] The present application also provides a device for quickly detecting burrs in cable connectors. Figure 7 , which shows a schematic structural diagram of a cable joint burr rapid detection device provided by an embodiment of the present invention, the device includes a memory 701, a processor 702, and a computer program 703 stored in the memory 701 and running on the processor 702, wherein when the processor 702 executes the computer program 703, the device can execute any of the cable joint burr rapid detection methods introduced above.

[0125] The embodiment of the present application also provides a computer program product. When the computer program product is run on a computer device, the computer device can execute any one of the cable joint burr rapid detection methods introduced above.

[0126] An embodiment of the present application also provides a computer-readable storage medium, in which a computer program code is stored. When the computer program code is executed on a computer device, the computer device can execute any one of the cable joint burr rapid detection methods introduced above.

[0127] In the embodiments provided in the present application, it should be understood that the provided computer device, computer program product and computer-readable storage medium are all used to execute the corresponding methods provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the methods provided above and will not be repeated here.

[0128] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0129] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for rapid detection of cable joint burrs, It is characterized in that The method comprises: Capturing the cable cross-section grayscale image of the cable cross-section at each shooting angle at the cable joint by a camera; acquiring all connected domains in the cable cross-section grayscale image; the shooting angles include a main viewing angle and at least two auxiliary viewing angles; In the cable cross-section grayscale image, the edge feature abnormality of each connected domain is determined according to the similarity between the shape of each connected domain and a circle and the consistency between the corner point density and the direction distribution of the corner points in the connected domain; According to the similarity in overall gray value, edge feature abnormality and position between each connected domain in the main viewing angle and all connected domains in each auxiliary viewing angle, the matching degree between each connected domain in the main viewing angle and each connected domain in each auxiliary viewing angle is determined; Determine the matching connected domains of each connected domain in the main viewing angle in each auxiliary viewing angle according to the matching degree; determine the abnormality degree of each connected domain in the main viewing angle according to the abnormality degree of edge features of each connected domain in the main viewing angle and the fluctuation of the matching degree of each corresponding matching connected domain; and screen out all abnormal burr areas in the main viewing angle according to the abnormality degree; Determine the possibility of the burr region of each connected domain under the main viewing angle according to the similarity between each connected domain under the main viewing angle and each abnormal burr region in grayscale gradient and grayscale and the corresponding edge feature abnormality; detect all real burr regions according to the burr region possibility and all abnormal burr regions; The method for obtaining the edge feature abnormality degree includes: Calculating the circularity of each connected domain in the cable cross-section grayscale image; Determine all corner points corresponding to each connected domain and the corner point direction of each corner point by using the corner point detection method; quantify the corner point direction of each corner point into a corner point direction angle according to the angle between the corner point direction of each corner point and the horizontal direction; calculate the corner point direction angle variance corresponding to all corner points in each connected domain; Determine the corner point density of each connected domain according to the number of corner points in each connected domain and the area of ​​the connected domain; Determine the edge feature abnormality of each connected domain according to the corner point direction angle variance, the corner point density and the circularity; the corner point direction angle variance and the circularity are both negatively correlated with the edge feature abnormality, and the corner point density is positively correlated with the edge feature abnormality; The method for obtaining the matching degree includes: Determine the pixel coordinates of the centroid point of each connected domain in the cable cross-section grayscale image and calculate the grayscale value mean of all pixels in each connected domain; take each connected domain under the main viewing angle as the target connected domain in turn; take each connected domain under each auxiliary viewing angle as the auxiliary connected domain in turn; Determining a corresponding reference grayscale difference according to a difference between a grayscale value mean of the target connected domain and a grayscale value mean of the auxiliary connected domain; Determining a corresponding reference position difference according to a Euclidean distance between pixel coordinates of a centroid point of the target connected domain and pixel coordinates of a centroid point of the auxiliary connected domain; Determining a corresponding edge anomaly difference according to a difference between an edge feature anomaly degree of the target connected domain and an edge feature anomaly degree of the auxiliary connected domain; The matching degree between the target connected domain and the auxiliary connected domain is determined according to the reference grayscale difference, the reference position difference and the edge anomaly difference; the reference grayscale difference, the reference position difference and the edge anomaly difference are all negatively correlated with the matching degree.

2. A method for rapid detection of cable joint burrs according to claim 1, It is characterized in that The method for obtaining the matching connected domain includes: The connected domain corresponding to the maximum matching degree between each auxiliary perspective and the target connected domain is used as the matching connected domain of the target connected domain in each auxiliary perspective.

3. A method for rapid detection of cable joint burrs according to claim 1, It is characterized in that The method for obtaining the abnormality degree includes: According to the matching degree variance between the target connected domain and all corresponding matching connected domains, the matching degree discreteness of the target connected domain is determined; according to the matching degree discreteness and the edge feature abnormality corresponding to the target connected domain, the abnormality degree of the target connected domain is determined; the matching degree discreteness and the edge feature abnormality corresponding to the target connected domain are both positively correlated with the abnormality degree.

4. A method for rapid detection of cable joint burrs according to claim 1, It is characterized in that The method for obtaining the abnormal burr area includes: Among all connected domains under the main viewing angle, the connected domains with abnormality greater than a preset abnormality threshold are regarded as abnormal burr regions.

5. A method for rapid detection of cable joint burrs according to claim 1, It is characterized in that The method for obtaining the possibility of the burr area includes: Determine the reference abnormal gray value according to the average gray value of all pixels in all abnormal burr areas; determine the reference abnormal gradient value according to the average gray gradient value of all edge pixels in all abnormal burr areas; The difference between the mean grayscale value of all pixels in each connected domain under the main viewing angle and the reference abnormal grayscale value is subjected to negative correlation mapping processing to determine the grayscale abnormality degree of each connected domain under the main viewing angle; The difference between the mean grayscale gradient value of all edge pixels in each connected domain under the main viewing angle and the reference abnormal gradient value is subjected to negative correlation mapping processing to determine the degree of gradient abnormality of each connected domain under the main viewing angle; According to the grayscale abnormality, the gradient abnormality and the edge feature abnormality corresponding to each connected domain under the main viewing angle, the possibility of the burr area of ​​each connected domain under the main viewing angle is determined; the edge feature abnormality corresponding to each connected domain under the main viewing angle, the grayscale abnormality and the gradient abnormality are all positively correlated with the possibility of the burr area.

6. A method for rapid detection of cable joint burrs according to claim 1, It is characterized in that The method for obtaining the real burr area includes: Among all connected domains under the main viewing angle, connected domains whose burr region probability is greater than a preset burr threshold and all abnormal burr regions are regarded as real burr regions.

7. A cable joint burr rapid detection system, It is characterized in that The system comprises: A data acquisition preprocessing module is used to acquire a grayscale image of a cable cross section at each shooting angle through a camera; and obtain all connected domains in the grayscale image of the cable cross section; the shooting angle includes a main viewing angle and at least two auxiliary viewing angles; A first determination module is used to determine the abnormality of the edge features of each connected domain in the cable cross-section grayscale image according to the similarity between the shape of each connected domain and a circle and the consistency of the corner point density and the direction distribution of the corner points in the connected domain; The method for obtaining the edge feature abnormality degree includes: Calculating the circularity of each connected domain in the cable cross-section grayscale image; Determine all corner points corresponding to each connected domain and the corner point direction of each corner point by using the corner point detection method; quantify the corner point direction of each corner point into a corner point direction angle according to the angle between the corner point direction of each corner point and the horizontal direction; calculate the corner point direction angle variance corresponding to all corner points in each connected domain; Determine the corner point density of each connected domain according to the number of corner points in each connected domain and the area of ​​the connected domain; Determine the edge feature abnormality of each connected domain according to the corner point direction angle variance, the corner point density and the circularity; the corner point direction angle variance and the circularity are both negatively correlated with the edge feature abnormality, and the corner point density is positively correlated with the edge feature abnormality; The second determination module is used to determine the matching degree between each connected domain under the main viewing angle and each connected domain under each auxiliary viewing angle according to the similarity in overall gray value, edge feature abnormality and position between each connected domain under the main viewing angle and all connected domains under each auxiliary viewing angle; The method for obtaining the matching degree includes: Determine the pixel coordinates of the centroid point of each connected domain in the cable cross-section grayscale image and calculate the grayscale value mean of all pixels in each connected domain; take each connected domain under the main viewing angle as the target connected domain in turn; take each connected domain under each auxiliary viewing angle as the auxiliary connected domain in turn; Determining a corresponding reference grayscale difference according to a difference between a grayscale value mean of the target connected domain and a grayscale value mean of the auxiliary connected domain; Determining a corresponding reference position difference according to a Euclidean distance between pixel coordinates of a centroid point of the target connected domain and pixel coordinates of a centroid point of the auxiliary connected domain; Determining a corresponding edge anomaly difference according to a difference between an edge feature anomaly degree of the target connected domain and an edge feature anomaly degree of the auxiliary connected domain; Determining the degree of matching between the target connected domain and the auxiliary connected domain according to the reference grayscale difference, the reference position difference and the edge anomaly difference; the reference grayscale difference, the reference position difference and the edge anomaly difference are all negatively correlated with the degree of matching; The abnormal burr area screening module is used to determine the matching connected domains of each connected domain in the main viewing angle in each auxiliary viewing angle according to the matching degree; determine the abnormal degree of each connected domain in the main viewing angle according to the abnormal degree of edge features of each connected domain in the main viewing angle and the fluctuation of the matching degree of each corresponding matching connected domain; and screen out all abnormal burr areas in the main viewing angle according to the abnormal degree; The real burr area detection module is used to determine the burr area possibility of each connected domain under the main viewing angle according to the similarity between each connected domain under the main viewing angle and each abnormal burr area in grayscale gradient and grayscale and the corresponding edge feature abnormality; all real burr areas are detected according to the burr area possibility and all abnormal burr areas.

8. A cable joint burr rapid detection device, It is characterized in that include: A memory and a processor; wherein the memory is used to store computer program code that can be executed on the processor; The processor is used to call and run the computer program code from the memory to implement the steps of a cable joint burr rapid detection method as described in any one of claims 1-6.

Images