Automatic inspection method for aerospace cable multi-type crimping connector contacts

Abstract

The invention discloses an automatic inspection method for multi-type crimping connector contacts of aerospace cables, which comprises the following steps of: proposing indentation texture gradient information clustering calculation, and determining the types of the crimping connector contacts; multi-direction pixel and weighted calculation are provided, the distance between a wire insulation layer and the edge of the wire pressing cylinder is obtained, and whether the distance is qualified or not is judged; profile screening detection and multi-area segmentation axis detection are provided, and whether the contact piece of the crimping connector is damaged, polluted and deformed or not is judged. The method is suitable for the application background of various varieties of the Uamil route beams, and the inspection method is more convenient, accurate and comprehensive.

Description

Technical Field

[0001] The invention relates to the field of aerospace and weapon equipment cable harness manufacturing, in particular to an automatic inspection method for contact pieces of various types of crimped connectors of aerospace cables. Background Art

[0002] During the assembly process of aerospace crimp connectors, the contact terminals and wires need to be connected by pressing, and then they are fixed in the insulation holes of the connectors to complete the assembly. During this assembly process, according to the requirements of standards such as QJ 2926-1997, the crimping quality of the crimping connector contacts needs to be 100% inspected. The inspection method is mainly through visual inspection of the appearance by the inspector. Usually, the crimping connector contacts include the wire barrel, observation hole, interference boss, and engagement parts, such as Figure 1 As shown in the figure, the criteria for determining qualified contacts are mainly:

[0003] (1) The conductor core must be clearly visible through the contact observation hole;

[0004] (2) The indentation position needs to be between the observation hole and the edge of the wire barrel;

[0005] (3) The distance between the wire insulation layer and the wire barrel is 0.2mm <a<1mm;

[0006] (4) The contacts of the crimped connector must not have any abnormal appearance such as damage, contamination, deformation, etc.

[0007] (5) The core wire of the conductor must not be leaking;

[0008] (6) The wire specifications match the wire barrel diameter.

[0009] Currently, the industry's consensus is to use computer vision to automate the inspection of crimp-type connector contacts, replacing manual labor with automated methods to address the inefficiency and error-prone nature of this process. However, aerospace crimp-type connectors are characterized by a wide variety of products and small batches. For example, the main series of aerospace crimp-type connectors include:

[0010] The current inspection methods and inspection processes only perform automated inspections on the characteristics of a single crimped contact; they do not match the production model of aerospace low-frequency cable networks, and cannot achieve compatible inspections of multiple quantities and varieties. The equipment utilization rate is low and cannot meet production needs; in addition, the existing technology can better solve the problem of automatic judgment for the visibility of the core wire in the observation hole, the absence of wire leakage in the conductor core wire, the position of the indentation, and the matching of the wire diameter, but lacks effective technical means for inspection points such as the distance between the conductor insulation layer and the wire barrel, and appearance abnormalities such as damage, contamination, and deformation. Summary of the Invention

[0011] The technical problem solved by the present invention is: to overcome the shortcomings of the existing technology, provide an automatic inspection method for contact parts of various types of crimp connectors of aerospace cables, realize compatible inspection of multiple quantities and varieties, and provide automated technical means for inspecting the distance and shape dimensions between the insulation layer of the wire and the crimping barrel.

[0012] The technical solution of the present invention is: an automatic inspection method for various types of crimp connector contacts of aerospace cables, wherein the crimp connector contacts include a wire barrel, an engaging member is installed on one side of the wire barrel, and a wire on the other side is fixed in the wire barrel, and the wire barrel has an indentation. The inspection method includes:

[0013] Acquire an image of a crimped connector contact, determine the indentation area, and use clustering calculation of the indentation texture gradient information to determine the type of the crimped connector contact.

[0014] After image segmentation, the de-interferenced images of the wire barrel and wire insulation layer are obtained. The images are adjusted to highlight the characteristics of the wire barrel and identify the edges of the wire barrel and wire insulation layer. For the wire core area between the edges, multi-directional pixel and weighted calculation are used to determine the distance between the wire insulation layer and the wire barrel edge, and then determine whether the distance is qualified.

[0015] After image processing and target extraction, the extraction results of the crimped connector contacts are obtained; through contour screening detection, it is determined whether the crimped connector contacts are damaged and contaminated; through multi-region segmented axis detection, it is determined whether the crimped connector contacts are deformed.

[0016] Furthermore, the indentation area is determined as follows:

[0017] The acquired image is denoised, and the denoised RGB image is converted into the HSV color space. The color space value difference between the crimped connector contact and the background in the image is used to screen and locate the indentation area.

[0018] Furthermore, for the indentation area, clustering calculation of indentation texture gradient information is used to determine the type of crimped connector contact. The specific method is as follows:

[0019] Calculating the gradient information of the indentation area to obtain a gradient information set A; the gradient information set A includes the gradient information of each coordinate point in the indentation area;

[0020] Find the coordinate point with the smallest x-axis coordinate in the gradient information set A, recorded as B0;

[0021] In a counterclockwise direction, calculate the vector B0B formed by the remaining points of the gradient information set and B0 n ; n ranges from 1 to N-1, where N is the number of coordinate points in the gradient information set A;

[0022] Calculating B0B n The angle between the vector and the y-axis is sorted from large to small to obtain an ordered set C of angle values;

[0023] Calculate the angle set C between the ordered set of angle values ​​and the standard image vector of the indentation of each crimped connector contact {C1, C2, ..., C m}, m ranges from 1 to M, M is the number of standard image vector angle sets; i indentation standard image vector angle sets with similarities between the ordered set C and the standard image vector angle set are screened, i < M, and a set K of angle values ​​with similarities above the preset range is obtained;

[0024] Eliminate the angle value set K from the ordered set C and the filtered indentation standard image vector angle set to form a new ordered set C' and the filtered indentation standard image vector angle set {C1', C2', ..., C i '};

[0025] Traverse the new ordered set C', calculate the angle value in the new ordered set C' in the angle set {C1', C2', ..., C i '}, the type corresponding to the angle set with the highest frequency is determined as the crimp connector contact type.

[0026] Furthermore, the image is adjusted to highlight the characteristics of the wire barrel. The specific method is as follows:

[0027] First, convert the image to grayscale, then adjust the image brightness to the lowest, and then increase the contrast to reduce the grayscale of the gray area and increase the grayscale of the dark area, thereby highlighting the characteristics of the wire barrel.

[0028] Further, identify the edges of the wire barrel and wire insulation by:

[0029] The vertical line of the edge of the wire barrel is identified by using Hough transform, and the vertical line of the edge of the wire insulation layer is identified by the pixel value difference between the wire insulation layer, the background and the wire core.

[0030] Furthermore, for the wire core area between the edges, multi-directional pixel and weighted calculation are used to obtain the distance between the wire insulation layer and the edge of the wire barrel. The specific method is as follows:

[0031] According to the pixel coordinates of the wire core area, a pixel set D is obtained;

[0032] Filter the pixel coordinates of the transition area between the conductor insulation layer and the conductor core;

[0033] According to the coordinate information in the pixel set D, the pixel set is divided into unit sets D1, D2, D3, D4, ... by vectors according to the direction of the fixed vector of the image;

[0034] Based on the image resolution, calculate the weighted sum of the pixel sizes in each unit set D1, D2, D3, D4, etc. The principle for determining the weight is: based on the pixel coordinates of the transition area obtained by screening, the closer the pixel coordinates are to the transition area, the smaller the weight;

[0035] The weighted sums obtained for each unit set are averaged to obtain the distance between the wire insulation and the edge of the barrel.

[0036] Furthermore, the image processing includes: performing image denoising, background removal, and grayscale binarization processing on the image.

[0037] Furthermore, the target extraction includes: performing color segmentation, boundary segmentation, and region segmentation on the pre-processed image, thereby obtaining a result of extracting the crimped connector contact.

[0038] Furthermore, the profile screening test is used to determine whether the crimped connector contacts are damaged or contaminated. The specific method is as follows:

[0039] Extracting a contour collection E of the crimped connector contact;

[0040] Add coordinate information to the contour set E to form a multidimensional matrix array F; calculate the areas of all contours in the contour set E to form a contour area set G; add the contour area set G to the multidimensional matrix array F to increase the area information in F;

[0041] If pixel points with an area less than 50 can be screened out in the multi-dimensional matrix array F, it is determined that the contact of the crimp connector is damaged and contaminated.

[0042] Furthermore, through multi-area segmented axis detection, it is determined whether the crimped connector contacts are deformed. The specific method is as follows:

[0043] The left and right sides of the wire barrel are defined as the left area and the right area;

[0044] The average value of the upper and lower edges of the left area and the right area is used as the point of the geometric horizontal axis of the area, and the set of all the horizontal axis points in the left area and the right area is obtained;

[0045] The two point sets are respectively fitted into two straight lines. If the angle between the two fitted straight lines is ≠180°, it is determined that the crimped connector contact is deformed.

[0046] The advantages of the present invention compared with the prior art are:

[0047] (1) The present invention proposes a clustering calculation method for indentation texture gradient information. Based on the indentation shape, it adaptively selects and matches crimped terminals of different indentation types. It locates the position of the indentation area in different assembly scenarios, replacing the existing method of only using tooling limits or camera field of view interception. The disadvantage of the existing technology is that it may be affected by the light source and may cause misidentification. It can only distinguish a single type of crimped contact. It has practical value in industries with relatively single batches, such as automotive wiring harnesses, but it is not suitable for the application background of aerospace wiring harnesses with a wide variety of types.

[0048] (2) Compared with the existing technology, the present invention determines the quality of contact crimping in multiple dimensions to ensure reliability. The present invention uses multi-directional pixel value summation calculation, contour screening detection, and multi-region segmented axis detection methods to check whether the crimped contact has abnormal wire insulation distance, excess contamination, abnormal bending, and other common unqualified conditions in the aerospace cable production process. Compared with the existing technology that can only perform feature recognition inspections on contact wire leakage, wire diameter, etc., the present invention is more accurate, comprehensive, and practical, and effectively solves the problem of inspecting unqualified crimped contacts caused by incorrect process parameters, tool errors, or improper operation in actual production, thereby improving equipment reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the contact structure of the crimping connector in the background technology;

[0050] Figure 2 Schematic diagram of the method of the present invention;

[0051] Figure 3 Schematic diagram of identifying the vertical line on the edge of the wire barrel of the present invention. DETAILED DESCRIPTION

[0052] In order to better understand the technical solution of the present invention, the specific implementation methods of the present invention are described below.

[0053] The present invention proposes an automatic inspection method for contact pieces of various types of crimped connectors for aerospace cables, the main steps of which are as follows: Figure 2 Shown, including:

[0054] Step 1: Determine the type of crimp connector contact

[0055] (1) Suppress the noise of the image by using existing image denoising methods such as median filtering;

[0056] (2) Convert the RGB image to the HSV color space, and use the color space value difference between the crimped connector contact and the background in the image to screen and locate the indentation area of ​​the crimped connector contact;

[0057] Specifically, the background area of ​​the indentation of the crimped connector contact has a color space value between [100, 43, 46] and [124, 255, 255] and is identified as the indentation area through screening;

[0058] (3) The type of contact of the crimped connector is determined by clustering and calculating the indentation texture gradient information. The specific method is as follows:

[0059] (3.1) Calculating the gradient information of the indentation area to obtain a gradient information set A; the gradient information set A includes the gradient information of each coordinate point in the indentation area;

[0060] (3.2) Find the coordinate point with the smallest x-axis coordinate in the gradient information set A, denoted as B0;

[0061] (3.3) Select the counterclockwise direction and calculate the vector B0B formed by the remaining points of the gradient information set and B0 n ; n ranges from 1 to N, where N is the number of coordinate points in the gradient information set A;

[0062] (3.4) Calculate B0B n The angle between the vector and the y-axis is sorted from large to small to obtain an ordered set C of angle values;

[0063] (3.5) Calculate the angle set C between the ordered set of angle values ​​and the standard image vector of the indentation of each crimped connector contact {C1, C2, ..., C m}, m ranges from 1 to M, M is the number of standard image vector angle sets; i indentation standard image vector angle sets with similarities between the ordered set C and the standard image vector angle set are screened, i < M, and a set K of angle values ​​with similarities above the preset range is obtained;

[0064] (3.6) The angle value set K is removed from the ordered set C and the filtered indentation standard image vector angle set to form a new set C' and the filtered indentation standard image vector angle set {C1', C2', ..., C i '};

[0065] (3.7) Traverse the new set C' and calculate the angle value in the new set C' in the angle set {C1', C2', ..., C i '}, the type corresponding to the angle set with the highest frequency is determined as the crimp connector contact type.

[0066] Step 2: Check the size of the crimped connector contacts

[0067] The method includes several main steps: image segmentation, image adjustment, vertical line recognition of the edge of the wire barrel, vertical line recognition of the edge of the wire insulation layer, size output, and size determination.

[0068] (1) Image segmentation: Segment the target image of the wire pressing cylinder to obtain the image of the wire pressing cylinder and the wire insulation layer after removing interference items such as the background.

[0069] (2) Image adjustment: First, convert the image to grayscale to avoid excessive color deviation caused by the reflection of the wire pressing cylinder under the light source. Then, adjust the brightness and contrast of the image, adjust the brightness of the picture to the lowest and increase the contrast, reduce the gray level of the gray area, increase the gray level of the dark area, and highlight the characteristics of the wire pressing cylinder.

[0070] (3) Identification of the perpendicular lines at the edge of the wire pressing cylinder: Use the Hough transform to screen and find the perpendicular lines at the edge of the wire pressing barrel. See Figure 3 as shown. The red line part in the figure is the perpendicular line at the edge part.

[0071] (4) Identification of the perpendicular lines at the edge of the wire insulation layer: The color of the wire insulation layer generally has a large difference in pixel values from other positions such as the background and the core wire. Locate the perpendicular lines at the edge of the wire insulation layer through pixel values.

[0072] (5) Dimension output: Obtain the distance between the wire insulation layer and the wire pressing cylinder through the multi-directional pixel sum and weighted value method proposed in this invention. Specifically:

[0073] (5.1) Arrange the coordinate positions of the wire core area to obtain the pixel set D.

[0074] (5.2) Screen the pixel coordinates in the transition area between the wire insulation layer and the wire core; The interface between the wire insulation layer and the wire core is not a flat cut, and this part of the detailed area is defined as the transition area.

[0075] (5.3) According to the coordinate information mapped in the pixel set D, in the fixed vector direction of the image, that is, the vector direction from the wire insulation layer to the wire pressing cylinder, divide the pixel set into unit sets D1, D2, D3, D4... with vectors.

[0076] (5.4) According to the image resolution, calculate the weighted sum of pixel sizes in each set of D1, D2, D3, D4...; The principle for determining the weight value is: According to the screened pixel coordinates, the closer the pixel coordinates are to the transition area, the smaller the weight value.

[0077] (5.5) Take the average value of the weighted sums obtained for each unit set, which is the distance d between the wire insulation layer and the edge of the wire pressing cylinder.

[0078] (6) Dimension determination: Determine the output distance dimension d between the wire insulation layer and the wire pressing cylinder. If 0.2mm < d < 1mm, it can be determined as qualified, otherwise unqualified.

[0079] Step 3: Inspection of the morphology of the contact of the crimping connector

[0080] It includes several main steps: image preprocessing, target extraction, and shape detection.

[0081] (1) Image preprocessing

[0082] Based on the characteristics of the crimped connector contacts, the image is processed in sequence: image denoising, background removal, and grayscale binarization.

[0083] (2) Target extraction

[0084] The preprocessed image is subjected to color segmentation, boundary segmentation, and region segmentation to obtain the results of crimped connector contact extraction.

[0085] (3) Morphology detection

[0086] Perform morphology inspection on the extracted results of crimped connector contacts, including:

[0087] (3.1) Determine whether the crimped connector contacts are damaged or contaminated:

[0088] Extracting a contour collection E of the crimped connector contact;

[0089] Add image relative coordinate information to the contour collection E to form a multidimensional matrix array F; calculate the areas of all contours in the contour collection E to form a contour area set G; add the contour area set G to the multidimensional matrix array F to increase the area information in F;

[0090] In the C array, based on production experience, points with an area less than 50 pixels are screened; the output of the screened contaminated areas, if the number is ≥1, is a defective product.

[0091] (3.2) Determine whether the crimped connector contacts are deformed:

[0092] For the binarized image, select the test area of ​​the test piece and define the left and right sides of the wire barrel as the left area and the right area. Specifically, in this embodiment, the left area of ​​the wire barrel is defined as the area 0.05l to 0.20l from the left edge of the image, and the right area of ​​the wire barrel is defined as the area 0.80l to 0.95l from the left edge of the image.

[0093] The average value of the upper and lower edges of the left area and the right area is used as the point of the geometric horizontal axis of the area, and the set of all the horizontal axis points in the left area and the right area is obtained;

[0094] Using the least squares method, the two point sets are fitted into two straight lines. If the angle between the two fitted straight lines is ≠180°, the crimped connector contact is judged to be unqualified.

[0095] It will be understood that the present invention is described by way of example, and it will be appreciated by those skilled in the art that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, under the teachings of the present invention, these features and embodiments may be modified to adapt to specific circumstances without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and any embodiment that falls within the scope of the claims of this application is intended to be within the scope of protection of the present invention.

[0096] The contents not described in detail in the specification of the present invention belong to the common knowledge of those skilled in the art.

Claims

1. A method for automatically inspecting contacts of various types of crimp connectors for aerospace cables, wherein the crimp connector contacts include a wire barrel, an engaging member mounted on one side of the wire barrel, a wire fixed in the wire barrel on the other side, and an indentation on the wire barrel, characterized in that: The inspection method comprises: Acquire an image of a crimped connector contact, determine the indentation area, and use clustering calculation of the indentation texture gradient information to determine the type of the crimped connector contact. After image segmentation, the de-interferenced images of the wire barrel and wire insulation layer are obtained. The images are adjusted to highlight the characteristics of the wire barrel and identify the edges of the wire barrel and wire insulation layer. For the wire core area between the edges, multi-directional pixel and weighted calculation are used to determine the distance between the wire insulation layer and the wire barrel edge, and then determine whether the distance is qualified. After image processing and target extraction, the extraction results of the crimped connector contacts are obtained; through contour screening detection, it is determined whether the crimped connector contacts are damaged and contaminated; through multi-region segmented axis detection, it is determined whether the crimped connector contacts are deformed.

2. The method for automatically inspecting contacts of various types of crimp connectors for aerospace cables according to claim 1, characterized in that: Determine the indentation area by: The acquired image is denoised, and the denoised RGB image is converted into the HSV color space. The color space value difference between the crimped connector contact and the background in the image is used to screen and locate the indentation area.

3. The method for automatically inspecting contacts of various types of crimped connectors for aerospace cables according to claim 1 or 2, characterized in that: For the indentation area, clustering calculation of indentation texture gradient information is used to determine the type of crimp connector contact. The specific method is as follows: Calculating the gradient information of the indentation area to obtain a gradient information set A; the gradient information set A includes the gradient information of each coordinate point in the indentation area; Find the coordinate point with the smallest x-axis coordinate in the gradient information set A, recorded as B0; In a counterclockwise direction, calculate the vector B0B formed by the remaining points of the gradient information set and B0 n ; n ranges from 1 to N-1, where N is the number of coordinate points in the gradient information set A; Calculating B0B n The angle between the vector and the y-axis is sorted from large to small to obtain an ordered set C of angle values; Calculate the angle value ordered set C and the angle set {C1, C2, ..., C m }, m ranges from 1 to M, M is the number of standard image vector angle sets; i indentation standard image vector angle sets with similarities between the ordered set C and the standard image vector angle set are screened, i < M, and a set K of angle values ​​with similarities above the preset range is obtained; Eliminate the angle value set K from the ordered set C and the filtered indentation standard image vector angle set to form a new ordered set C' and the filtered indentation standard image vector angle set {C1', C2', ..., C i '}; Traverse the new ordered set C', calculate the angle value in the new ordered set C' in the angle set {C1', C2', ..., C i '}, the type corresponding to the angle set with the highest frequency is determined as the crimp connector contact type.

4. The method for automatically inspecting contacts of various types of crimp connectors for aerospace cables according to claim 1, characterized in that: Adjust the image to highlight the characteristics of the wire barrel. The specific method is: First, convert the image to grayscale, then adjust the image brightness to the lowest, and then increase the contrast to reduce the grayscale of the gray area and increase the grayscale of the dark area, thereby highlighting the characteristics of the wire barrel.

5. The method for automatically inspecting contacts of various types of crimp connectors for aerospace cables according to claim 1, characterized in that: Identify the edges of the wire barrel and wire insulation by: The vertical line of the edge of the wire barrel is identified by using Hough transform, and the vertical line of the edge of the wire insulation layer is identified by the pixel value difference between the wire insulation layer, the background and the wire core.

6. The method for automatically inspecting contacts of various types of crimp connectors for aerospace cables according to claim 1, 4 or 5, characterized in that: For the wire core area between the edges, multi-directional pixel and weighted calculation are used to obtain the distance between the wire insulation layer and the edge of the wire barrel. The specific method is as follows: According to the pixel coordinates of the wire core area, a pixel set D is obtained; Filter the pixel coordinates of the transition area between the conductor insulation layer and the conductor core; According to the coordinate information in the pixel set D, the pixel set is divided into unit sets D1, D2, D3, D4, ... by vectors according to the direction of the fixed vector of the image; Based on the image resolution, calculate the weighted sum of the pixel sizes in each unit set D1, D2, D3, D4, etc. The principle for determining the weight is: based on the pixel coordinates of the transition area obtained by screening, the closer the pixel coordinates are to the transition area, the smaller the weight; The weighted sums obtained for each unit set are averaged to obtain the distance between the wire insulation and the edge of the barrel.

7. The method for automatically inspecting contacts of various types of crimp connectors for aerospace cables according to claim 1, characterized in that: The image processing includes: performing image denoising, background removal, and grayscale binarization processing on the image.

8. The method for automatically inspecting contacts of various types of crimp connectors for aerospace cables according to claim 1, characterized in that: The target extraction includes: performing color segmentation, boundary segmentation, and region segmentation on the pre-processed image, thereby obtaining a result of extracting the contact of the crimping connector.

9. The method for automatically inspecting contacts of various types of crimp connectors for aerospace cables according to claim 1, 7 or 8, characterized in that: Through profile screening testing, the crimped connector contacts are judged to be free of damage and contamination. The specific methods are as follows: Extracting a contour collection E of the crimped connector contact; Add coordinate information to the contour set E to form a multidimensional matrix array F; calculate the areas of all contours in the contour set E to form a contour area set G; add the contour area set G to the multidimensional matrix array F to increase the area information in F; If pixel points with an area less than 50 can be screened out in the multi-dimensional matrix array F, it is determined that the contact of the crimp connector is damaged and contaminated.

10. The method for automatically inspecting contacts of various types of crimp connectors for aerospace cables according to claim 1, 7 or 8, characterized in that: Through multi-area segmented axis detection, determine whether the crimped connector contacts are deformed. The specific method is as follows: The left and right sides of the wire barrel are defined as the left area and the right area; The average value of the upper and lower edges of the left area and the right area is used as the point of the geometric horizontal axis of the area, and the set of all the horizontal axis points in the left area and the right area is obtained; The two point sets are respectively fitted into two straight lines. If the angle between the two fitted straight lines is ≠180°, it is determined that the crimped connector contact is deformed.

Images