A wound cable skin detection device and detection method thereof

Abstract

A winding cable skin detection device and a detection method thereof belong to the field of cable laying and cable skin damage detection. In the existing cable laying process, there is no function to detect the skin of the cable, and the cable may malfunction due to skin damage during subsequent use. A winding cable skin detection device, the bottom bracket is located at the bottom of the winding cable skin detection device, and the carrying box is fixed above the bottom bracket; on the side of the carrying box, from front to back, a winding system, a wire system, an auxiliary bracket, a detection system, a speed measurement system, another auxiliary bracket and another winding system are installed in sequence, and the winding ends of the wire system and the two winding systems are exposed to the outside of the carrying box; a touch system and control buttons are embedded and installed on the top surface of the carrying box. The detection result of whether there is an abnormality in the cable skin of the present invention is accurate.

Description

Technical Field

[0001] The invention relates to a winding cable skin detection device and a detection method thereof. Background Art

[0002] During the laying process of some cables, the cable surface is not inspected, and the cables may malfunction due to damaged surface during subsequent use. Summary of the Invention

[0003] The purpose of the present invention is to solve the problem that the existing cable laying process has no function of detecting the cable skin, and to propose a winding cable skin detection device and a detection method thereof.

[0004] A winding cable skin detection device, comprising: a bottom bracket, a carrying box, a winding system, a touch system, a control button, a wire system, a detection system, a speed measurement system, and a set of auxiliary brackets;

[0005] The bottom bracket is located at the lowest end of the wound cable skin detection device, and the carrying box is fixed above the bottom bracket; on one side of the carrying box, a winding system, a wire system, an auxiliary bracket, a detection system, a speed measurement system, another auxiliary bracket, and another winding system are installed in sequence from front to back, and the installation ends and power source of the wire system and the two winding systems are located inside the carrying box, while the winding ends of the wire system and the two winding systems are exposed outside the carrying box; a touch system and control buttons are embedded on the top surface of the carrying box;

[0006] in,

[0007] The touch control system is used to set the operating parameters of the entire device and monitor the detection status of the cable; the touch control system integrates the control unit of the winding cable skin detection device, and the touch control system and the control button portion with the control unit are located inside the carrying box, while the operating portion of the touch control system and the control button is exposed outside the carrying box;

[0008] The control button is used to start, stop and emergency stop the winding cable skin detection device;

[0009] The two winding systems, the two wire systems, the detection system, the speed measurement system and the control buttons are simultaneously connected to the control unit of the touch control system.

[0010] A method for detecting the surface of a wound cable, the method comprising the following steps:

[0011] Step 1: The control system controls the operation of the motors of the winding system and the conductor system, and at the same time controls the detection camera of the detection system to capture the cable surface image in real time;

[0012] Step 2: Use image detection algorithms to identify and process images taken from multiple angles. If damage is detected, the motors of the winding and conductor systems are stopped, and images of the detected damage are displayed to the user. The user confirms whether it is wire damage. After the user confirms, the control system continues to operate. If not, detection continues.

[0013] Step 3: The control system simultaneously collects information from the speed sensor in real time to detect whether the current take-up speed is greater than or less than the pay-out speed. When the speed difference exceeds a preset threshold, the control system will control the corresponding motor to stop running. When the speed difference is reduced to within the range of the preset value, the control system will control the stopped motor to continue running. When the entire cable reel is tested, the system stops testing and controls the motor to stop.

[0014] The beneficial effects of the present invention are:

[0015] The present invention has a compact structure and a simple and fast detection workflow.

[0016] The method of the present invention is to perform edge detection on the wire image after threshold processing, and then accurately search the upper and lower edges of the wire to obtain the upper and lower edge point sets. For normal wires, they have the characteristics of stable width, regular color, and uniform and smooth curvature of the upper and lower edges. Therefore, by calculating the average width of the upper and lower edge point sets and then comparing them, most damage problems can be found. At the same time, the slope and convexity of the upper and lower edge point sets are partitioned and calculated, and abnormal changes in slope and convexity are found to find bending anomalies. The wire is a curved strip area in the figure. In order to facilitate processing, for problems existing inside, the algorithm integrates the curved strip area into a rectangular long strip area, and then detects abnormal small areas (blocks) in the rectangular area based on thresholds, thereby finding abnormalities inside the wire. By ensuring that each X position has an edge point, accurate edge detection steps, and accurate and complete detection of all points on the upper and lower edges, the cable skin detection effect of the method of the present invention is accurate and can effectively detect outer skin damage. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 The present invention relates to a winding cable skin detection device;

[0018] Figure 2 This is a rear side view of the winding cable skin detection device according to the present invention;

[0019] Figure 3 This is a partial front view of the winding cable skin detection device according to the present invention;

[0020] Figure 4 The present invention relates to a winding system;

[0021] Figure 5 An exploded diagram of the winding system involved in the present invention;

[0022] Figure 6 A cross-sectional view of a wire system according to the present invention;

[0023] Figure 7 This is an independent view of the wire system involved in the present invention;

[0024] Figure 8 An exploded diagram of the conductor system involved in the present invention;

[0025] Figure 9 This is an installation location diagram of the detection system involved in the present invention;

[0026] Figure 10 The detection system involved in the present invention;

[0027] Figure 11 This is a top view of the detection system involved in the present invention;

[0028] Figure 12 This is an exploded diagram of the detection system involved in the present invention;

[0029] Figure 13 The auxiliary bracket of the present invention;

[0030] Figure 14 This is a working diagram of the winding cable skin detection device according to the present invention;

[0031] Figure 15 This is a working principle diagram of the conductor system involved in the present invention;

[0032] Figure 16 A flow chart of the method according to the present invention;

[0033] Figure 17 The wire image of the present invention;

[0034] Figure 18 The wire image of the present invention;

[0035] Figure 19 This is a schematic diagram of the steps of locating left and right endpoints in the wire image of the present invention;

[0036] Figure 20 This is a schematic diagram of the steps of locating left and right endpoints in the wire image of the present invention;

[0037] Figure 21 An illustration of Canny wire edge detection in a wire image according to the present invention;

[0038] Figure 22 An illustration of Canny wire edge detection in a wire image according to the present invention;

[0039] Figure 23 Schematic diagram of the wire edge detection step in the wire image according to the present invention;

[0040] Figure 24 The figure is a schematic diagram of the wire edge detection step in the wire image according to the present invention. DETAILED DESCRIPTION Specific implementation method one:

[0042] A winding cable skin detection device in this embodiment, such as Figure 1 As shown, its components include: bottom bracket 1, carrying box 2, winding system 3, touch system 4, control button 5, wire system 6, detection system 7, speed measurement system 8, and a set of auxiliary brackets 9;

[0043] The bottom bracket 1 is located at the lowest end of the winding cable skin detection device, and the carrying box 2 is fixed above the bottom bracket 1, and the bottom bracket 1 supports the carrying box 2. On one side of the carrying box 2, a winding system 3, a wire system 6, an auxiliary bracket 9, a detection system 7, a speed measurement system 8, another auxiliary bracket 9 and another winding system 3 are installed in sequence from front to back, and the installation ends and power source parts of the wire system 6 and the two winding systems 3 are located inside the carrying box 2, and the winding ends of the wire system 6 and the two winding systems 3 are exposed outside the carrying box 2 for operations such as winding and guiding the cables. A touch system 4 and a control button 5 are embedded and installed on the top surface of the carrying box 2 from front to back.

[0044] in,

[0045] The touch control system 4 is used to set the operating parameters of the entire device and monitor the detection status of the cable. The touch control system 4 integrates the control unit of the winding cable skin detection device, which is used to collect and control the working status of the entire device. The touch control system 4 and the control button 5 are located inside the carrying box 2, and the operating parts of the touch control system 4 and the control button 5 are exposed outside the carrying box 2.

[0046] The control button 5 is located on one side of the touch control system 4 and is used to start, stop, and emergency stop the winding cable skin detection device.

[0047] The two winding systems 3 , the two wire systems 6 , the detection system 7 , the speed measurement system 8 and the control button 5 are simultaneously connected to the control unit of the touch control system 4 ; the control unit is the CPU. Specific implementation method two:

[0049] Different from the first embodiment, this embodiment is a winding cable skin detection device.

[0050] The installation end of the winding system 3 also includes a winding fixing seat 3a and a winding motor bracket 3b;

[0051] The power source part of the winding system 3 also includes a winding motor 3c and a winding coupling 3d;

[0052] The winding end of the winding system 3 further includes a winding bearing seat 3e, a winding shaft 3f and a winding wheel 3g;

[0053] The winding fixing seat 3a is the fixing and supporting component of the winding system 3 and is installed on the inner bottom surface of the carrying box 2. The winding motor bracket 3b and the winding bearing seat 3e are arranged and fixed on the top surface of the winding fixing seat 3a; the winding motor 3c is fixed to the winding motor bracket 3b, and the rear end of the winding shaft 3f passes through the winding bearing seat 3e and is connected to the output shaft of the winding motor 3c through the winding coupling 3d; the front end of the winding shaft 3f is installed with a winding wheel 3g; and the front end of the winding shaft 3f passes through the side wall of the carrying box 2, and the winding wheel 3g is located on the outside of the carrying box 2;

[0054] The output shaft of the winding motor 3c drives the winding wheel 3g to rotate together through the winding coupling 3d and the winding shaft 3f;

[0055] The bearing seat 3e plays an auxiliary fixing role for the winding shaft 3f, ensuring that the winding shaft 3f and the winding wheel 3g rotate smoothly during operation. Specific implementation method three:

[0057] Different from the first or second embodiment, in this embodiment, a winding cable skin detection device, the installation end of the wire system 6 further includes a wire fixing seat 6a, a wire motor bracket 6b

[0058] The power source part of the conductor system 6 also includes a conductor motor 6c and a conductor coupling 6d;

[0059] The winding end of the conductor system 6 further includes a conductor bearing seat 6e, a conductor wheel bracket 6f, a conductor wheel 6g, a conductor sliding block 6h, a conductor screw shaft 6i, a conductor support 6j, a conductor sliding shaft 6k, a conductor support plate 6l, and a conductor lead screw nut 6m;

[0060] The wire fixing seat 6a is a fixing and supporting component of the wire system 6, which is installed on the inner bottom surface of the bearing box 2. The wire motor bracket 6b and the wire bearing seat 6e are arranged and fixed on the top surface of the wire fixing seat 6a in sequence; the wire motor 6c is fixed on the wire motor bracket 6b, and the rear end of the wire screw shaft 6i passes through the wire bearing seat 6e and is connected to the output shaft of the wire motor 6c through the wire coupling 6d; the front end of the wire screw shaft 6i is provided with a screw structure, and the front end of the wire screw shaft 6i passes through the side wall of the bearing box 2 and extends to the outside of the bearing box 2; the wire support plate 6l is fixed to the outside of the bearing box 2, and the two wire supports 6j are fixed to the wire support A wire sliding shaft 6k is mounted between two wire supports 6j at the front and rear ends of the upper surface of the plate 61; a wire sliding block 6h is mounted on the wire sliding shaft 6k and can slide back and forth along the wire sliding shaft 6k; a wire lead screw nut 6m is embedded in the wire sliding block 6h; the top of the wire sliding block 6h is connected to a wire pulley bracket 6f, and three freely rotatable wire pulleys 6g are mounted on the wire pulley bracket 6f; a portion of the wire screw shaft 6i outside the carrying box 2 passes through the wire support 6j, the wire sliding block 6h, and another wire support 6j in sequence, and forms a lead screw nut pair with the wire lead screw nut 6m embedded in the wire sliding block 6h;

[0061] The output shaft of the wire motor 6c drives the wire screw shaft 6i to rotate together through the wire coupling 6d; since the wire screw shaft 6i is connected to the wire lead screw nut 6m in the form of a lead screw nut pair, and the wire lead screw nut 6m is installed and fixed on the wire sliding block 6h, the wire sliding block 6h slides back and forth along the wire sliding shaft 6k under the action of the lead screw nut pair, thereby driving the wire wheel bracket 6f and the wire wheel 6g installed on the wire wheel bracket 6f to move back and forth together. Specific implementation method four:

[0063] Different from the third embodiment, this embodiment is a winding cable skin detection device.

[0064] The detection system 7 also includes a detection system base 7a, a camera bracket 7b, a set of detection cameras 7c, a detection box 7d, a detection end cover 7e, and a detection system mounting frame 7f;

[0065] The detection system mounting frame 7f is fixed to the outside of the carrying box 2, the detection system base 7a is fixed on the detection system mounting frame 7f, and the detection box 7d is fixed to the top of the detection system base 7a; the detection box 7d is a hollow structure, and a group of mounting holes are distributed on the side wall of the box body of the detection box 7d, and the mounting holes are distributed in sequence at equal intervals along the circumferential direction of the cross section of the detection box 7d; each detection camera 7c is fixed to the side wall of the detection box 7d through a camera bracket 7b, and the shooting end of the detection camera 7c is respectively inserted into the detection box 7d through the mounting hole; detection end covers 7e are installed at both ends of the detection box 7d, and the panel of the detection end cover 7e is provided with a notch to facilitate the passage of the cables to be detected. Specific implementation method five:

[0067] What is different from the fourth embodiment is that in this embodiment of a winding cable skin detection device, the number of the detection cameras 7c is 3. Specific implementation method six:

[0069] Different from the first, second, fourth or fifth embodiments, this embodiment is a winding cable skin detection device.

[0070] The auxiliary bracket 9 also includes an auxiliary rod 9a and an auxiliary wheel 9b. The two auxiliary brackets 9 are fixed on both sides of the detection system 7 on the outer side wall of the carrying box 2; the auxiliary rod 9a is L-shaped, and the end of one right-angled side of the auxiliary rod 9a is fixed on the outer side wall of the carrying box 2, and the end of the other right-angled side of the auxiliary rod 9a serves as a hanging end, and a rotating shaft is set on the hanging end. A freely rotatable auxiliary wheel 9b is installed on the rotating shaft of the auxiliary rod 9a, and the outer edge surface of the auxiliary wheel 9b has a groove. Specific implementation method seven:

[0072] Different from the sixth embodiment, in this embodiment, a winding cable skin detection device is provided, and the speed measuring system 8 is installed on the outside of the carrying box 2, between the auxiliary bracket 9 and the detection system 7, to measure the movement speed of the passing cable. Specific implementation method eight:

[0074] A method for detecting the surface of a wound cable in this embodiment is as follows: Figure 16 As shown, the detection method comprises the following steps:

[0075] Step 1: The control system controls the operation of the motors of the winding system (3) and the conductor system (6), and simultaneously controls the detection camera (7c) of the detection system (7) to capture a picture of the cable surface in real time;

[0076] Step 2: Use an image detection algorithm to identify and process the pictures taken from multiple angles. If damage to the skin is detected, the motors of the winding system (3) and the wire system (6) are controlled to stop running, and the pictures of the detected damage are displayed to the user. The user confirms whether it is line damage. After the user confirms, the control system continues to run; if not, the detection continues;

[0077] Step 3: The control system simultaneously collects information from the speed sensor in real time to detect whether the current take-up speed is greater than or less than the pay-out speed. When the speed difference exceeds a preset threshold, the control system will control the corresponding motor to stop running. When the speed difference is reduced to within the range of the preset value, the control system will control the stopped motor to continue running. When the entire cable reel is tested, the system stops testing and controls the motor to stop. Specific implementation method nine:

[0079] Different from the eighth embodiment, in the present embodiment, a method for detecting the skin of a wound cable is used. In step 2, the process of using an image detection algorithm to identify and process images taken from multiple angles is as follows:

[0080] Step 21: Image preprocessing:

[0081] 1.1) Read an original wire image; for example Figure 17 and Figure 18 Images of wires shown respectively;

[0082] 1.2) Normalize the size of the wire image and convert it into a single-channel grayscale image;

[0083] 1.3) Using the reduced image, the area where the wires are located is preliminarily located based on a fixed threshold;

[0084] 1.4) Apply low threshold processing to the area formed by the wires in the image, and apply high threshold processing to the area formed by the background to remove interference within the background;

[0085] 1.5) Use high-pass filtering to smooth the data and use mean filtering to remove noise;

[0086] Step 22: For the wire image after threshold processing, accurately search for the upper and lower edges of the wire to obtain the upper and lower edge point sets of the wire, thereby detecting the upper and lower edges of the wire;

[0087] Step 23: Detect abnormality at the edge of the wire;

[0088] Step 24: Regularize the area where the wire is bent into a rectangular area, and then detect blocks with abnormal strength in the area to detect abnormal conditions inside the wire. Specific implementation method ten:

[0090] Different from the ninth embodiment, in the present embodiment, a method for detecting the skin of a wound cable is described. In step 22, for the threshold-processed wire image, the upper and lower edges of the wire are accurately searched to obtain the upper and lower edge point sets of the wire, thereby detecting the upper and lower edges of the wire. The process includes positioning the left and right endpoints of the wire and searching for the wire edge, specifically:

[0091] Step 221: The process of locating the left and right end points of the wire is as follows:

[0092] First, expand the left and right edges of the image to provide a buffer zone. Assume that the wires in the image are placed horizontally and start and end at the left and right edges.

[0093] After that, we mask most of the middle area of ​​the image and only extract the left and right narrow areas of the image. This way, we can significantly increase the running speed by only processing the narrow areas.

[0094] Afterwards, the left and right regions were binarized using a fixed threshold;

[0095] Afterwards, the Shi-Tomasi corner detection is used to extract the four strongest endpoints in the left and right regions. The Shi-Tomasi corner detection includes the following steps:

[0096] (a) The Shi-Tomasi method is used to calculate the grayscale value change of each pixel in the left and right regions. The calculation method is:

[0097]

[0098] Where E(u, v) is the window grayscale value, Ix and Iy are the partial derivatives of the grayscale image in the x and y directions, and w(x, y) is the window function;

[0099] (b) The matrix form of E(u, v) is:

[0100] Where M is the autocorrelation matrix, which is:

[0101] (c) Calculate the eigenvalues ​​λ1 and λ2 of the M matrix;

[0102] (d) Determine the corner points based on the matrix eigenvalues, λ1≥λ max And λ2≥λ max , where λ max is the corner detection threshold;

[0103] After that, the positions of the four endpoints are compared in the left and right regions, and two extreme points are selected to obtain the four rough coordinates of the wire at the left and right boundaries of the image, as shown in the figure. Figure 19 and 20The upper and lower edges of the left and right ends of the middle wire shown and the boundary intersection points;

[0104] Step Two Two Two, the process of the wire edge search part described above is specifically as follows:

[0105] First, perform Canny edge detection using a low threshold, Figure 21 and Figure 22 As shown in the figure of Canny wire edge detection, ensure that more edge information is retained

[0106] (a), Perform a convolution operation on the wire image using a convolution operator, * represents the convolution operation, and obtain its gradient value:

[0107]

[0108] The magnitude G and angle θ of the gradient are calculated by the following method:

[0109]

[0110] (b), After obtaining the gradient magnitude and angle of the wire image, for any pixel point A in the image, if the pixel value of A is not the maximum among the pixel values in its neighborhood, then A is not an edge point of the image; if the pixel value of A is the maximum among the pixel values in its neighborhood, then A is an edge point of the image;

[0111] (c), Define the high threshold H and low threshold L of the image. If the gradient G of the pixel point > H, then the pixel point is a true edge; if the gradient G of the pixel point < L, then the pixel point is a non-edge; when the gradient G of the pixel point is between L and H, judge whether there are true edge points in its surrounding neighborhood. If so, consider this point as a true edge point, otherwise consider this point as a non-edge point;

[0112] After that, on the binary edge image, re-locate and search for the left and right endpoints of the wire to obtain 4 accurate coordinates;

[0113] After that, from the left endpoint to the right endpoint direction, perform a point-by-point accurate search for the upper and lower edges, save the searched points in an array, and use the boundary tracking method to detect the upper and lower edges of the wire; among them, the boundary tracking method adopts the following two-level logic:

[0114] Level 1: Search for the next edge point within the set distance range in the order of the specified direction priority, and repeat continuously; if the distance range is exceeded, usually there is an obvious problem with the edge, and perform exception handling;

[0115] Level 2: Determine whether the searched edge point is adjacent to the previous point. If not, there is a gap in the middle. Search for edge points point by point in the reverse priority order from the searched edge point. If no edge point exists, fill in the points in a step-by-step manner until the X position of the previous point is reached.

[0116] After that, through point-by-point search, all points on the upper and lower edges are finally detected accurately and completely. If the position of the right endpoint is not finally found, exception processing is performed. Specific implementation method eleven:

[0118] Different from the tenth embodiment, in the present embodiment, a method for detecting the skin of a wound cable, the boundary tracking method may also adopt a boundary tracking method with a third-level logic. The third-level logic is:

[0119] Level 3: If the reverse connection is correct, it will be adjacent to the previous point in the Y direction. If it is not adjacent, it means that there is an error in the reverse connection. All reverse connection points need to be traced back one by one towards the power edge to ensure an uninterrupted connection.

[0120] In reality, the edges of ordinary wires are relatively clear. If the reverse edge search fails to successfully connect to the previous edge point, it indicates a discontinuity in the Y direction. A difference of 1 or less indicates no discontinuity. A noticeable anomaly in the reverse connection usually indicates the presence of foreign matter or damage. Therefore, a simple approach currently used is to directly output an edge anomaly when the Y-direction discontinuity exceeds a certain threshold. The threshold range is 1-5. To reduce false alarms, it can be set to 5 depending on the actual situation. Specific implementation method 12:

[0122] Different from the eleventh embodiment, in the present embodiment, a method for detecting the surface of a wound cable is used. In steps two and three, the process of detecting abnormalities on the edge of the wire is as follows:

[0123] Step 231: Interval sampling: calculate the height difference between adjacent points every 10 points, which represents the slope change;

[0124] Assume that point Vp1 (x1, y1) and point Vp2 (x2, y2) are separated by 10 points. Calculate K_Vp1 = y2 - y1. K_Vp1 represents the slope change of point Vp1.

[0125] Step 232: For every three consecutive sampling points, calculate the height difference between the middle point and the midpoints of the lines connecting the two sides, which represents the concavity:

[0126] Assume that point Vp3 (x3, y3), point Vp4 (x4, y4) and point Vp5 (x5, y5) are three consecutive sampling points, M_Vp4 represents the concavity and convexity of point Vp4, M_Vp4 = (y3 + y5) / 2 - y4;

[0127] Step 2, 3, 3. Smooth the slope and convexity of the wire to prevent repeated changes and sudden abnormally high values. By setting threshold rules, determine whether the wire has a bend problem:

[0128] Set the slope change threshold KK and the concavity threshold MM. For any pixel point Vp, when K_Vp>KK or M_Vp>MM, it is determined that the wire has a bending problem.

[0129] Steps 2, 3, and 4: Since there may be anomalies in the narrow areas on the left and right sides of the image, mask the edge points of the narrow areas on the left and right sides of the image and calculate the average width between the upper and lower edges of the middle area, as well as the maximum and minimum widths at each position:

[0130] Mask the edge points of the left and right narrow sides of the image. Assume that the upper and lower edge points of any position in the middle area are points Vp6 (x6, y6) and Vp7 (x7, y7). Therefore, calculate the width of the position width = y7-y6. After calculating the width of all the positions in the middle area, we get the width value set {width1, width2, ...width n}, where the maximum and minimum values ​​are denoted as max_width and min_width;

[0131] Average width

[0132] Steps 2, 3, and 5 compare the maximum and minimum widths with the average value to determine whether the wire width is abnormal:

[0133] Normal wires do not change in width. Comparing the maximum and minimum widths with the average value can be used to determine whether the wire has width abnormalities. This method is quite effective and can effectively detect sheath damage. Assuming the width face threshold WW, if |max_width-aver_width|>WW or |min_width-aver_width|>WW, it can be determined that the wire has width abnormalities. Figure 23 and Figure 24 Wire edge detection results are performed in the wire image shown. Specific implementation method thirteen:

[0135] Different from the twelfth embodiment, in the present embodiment, a method for detecting the surface of a wound cable is used. In step 24, the bent area of ​​the wire is regularized into a rectangular area, and then blocks with abnormal strength are detected within the area, thereby detecting abnormal conditions inside the wire. Specifically, the process is as follows:

[0136] Step 241: Assume that the wires in the image are arranged in a horizontal direction. For the image, the X-axis direction is the horizontal direction and the Y-axis direction is the vertical direction.

[0137] The sampling nearest neighbor interpolation method is used to normalize the vertical lines inside the wires in each X direction to the average width, and finally the wire area is extracted as a rectangular area;

[0138] Step 242: Set the width values ​​near the upper and lower edges of the masked rectangular area to eliminate the influence of the outer edges on the inner part;

[0139] Step 243: In the normal area inside the wire, the intensity changes little along the wire (there is a dim change in the vertical direction), so the average intensity is calculated at each Y position in the rectangular area:

[0140] In the rectangular area, any point along the line is denoted as Set_j={p 1j (x1,j),p 2j (x2,j),…,p nj (x n ,j)}, where p nj (x n ,j) represents pixel p nj The coordinates of (x n ,j);I(p nj ) represents the intensity value of the point;

[0141] Calculate average intensity Calculate all j in the vertical direction within the rectangular area;

[0142] Step 244: Compare each point within the rectangular area with the average intensity of its Y position. If the change exceeds the set threshold, it will be recorded as an internal outlier:

[0143] Set the intensity threshold TT. For any point Vp(x, y) in the rectangle, when |I(Vp)-intensity_y|>TT, mark the point as an outlier.

[0144] Steps 2, 4, and 5: Agglomerate adjacent abnormal points into zones, use contour extraction to determine the range of the abnormal zone, and calculate its area. When the internal abnormality exceeds a certain area, it indicates that there is a patch of abnormal strength inside the wire:

[0145] Set the area threshold AA, aggregate adjacent abnormal points into an area, and count the number of points in the abnormal area QQ. When QQ>AA, it is determined that there is a patch with abnormal strength inside the wire.

[0146] Working principle:

[0147] Before use, the winding cable skin detection device must first be used to wind the cable to be tested on the winding reel 3g. After winding, the winding reel 3g of the cable to be tested is installed on the winding system 3 at the rear end, here referred to as the winding reel to be tested 3g1; then an empty winding reel 3g is installed on the winding system 3 at the front end, here referred to as the tested winding reel 3g2.

[0148] Remove one end of the cable 10 to be tested from the reel 3g1 to be tested and insert it into the testing system 7 via the auxiliary wheel 9b of the auxiliary bracket 9 near the rear end. The cable 10 then passes through the U-shaped notch of the testing end cap 7e on one side and into the testing box 7d. It then exits the U-shaped notch of the testing end cap 7e on the other side and passes through the auxiliary wheel 9b of the auxiliary bracket 9 near the front end, around the three wire reels 6g of the wire system 6, before being secured to the tested reel 3g2. Adjust the two auxiliary brackets 9 to the appropriate position so that the cable 10 precisely enters and exits the U-shaped notch of the testing end cap 7e at both ends, ensuring smooth operation of the cable 10 during testing.

[0149] The operator needs to perform manual inspection on the cable 10 that has been pulled out from the winding reel 3g1 to be tested, because the cable 10 that has passed through the detection system 7 during the threading stage cannot be inspected by the detection system 7 again. Manual inspection can avoid quality problems on the surface of this section of cable 10.

[0150] According to the color and diameter of the cable 10, relevant detection parameters such as the corresponding attributes of the rotation speed of the front and rear winding wheels and the thresholds for detection judgment are set, and then the start button in the control button 5 is pressed to start the detection process of the cable 10.

[0151] When the cable 10 to be inspected passes through the inspection box 7d, three inspection cameras 7c arranged in sequence with a circumferential interval of 120° will capture images of the surface of the cable 10 from three directions. The collected data is transmitted to the control unit located in the touch system 4, and an algorithm program is used to determine whether there are defects in the cable surface.

[0152] The speed measuring system 8 detects the running speed of the cable 10. The cable 10 is pulled by the front-end measured reel 3g2 and rotates in conjunction with the rear-end measured reel 3g1. Therefore, it is necessary to measure the running speed of the measured cable 10 in real time so as to adjust the rotation speed of the corresponding conductor motor 6c to match the speeds of the front and rear reels, thereby preventing the measured cable 10 from being too loose in the detection system 7 and affecting the detection effect, or from being too tight and causing the cable to break.

[0153] The wire system 6 is used to ensure that the cable 10 is evenly wound on the measured winding wheel 3g2. The wire motor 6c drives the wire sliding block 6h to move through the screw nut pair composed of the wire screw shaft 6i and the wire screw nut 6m. The forward and reverse rotation of the wire motor 6c will be converted into the reciprocating motion of the wire sliding block 6h, thereby making the cable 10 evenly wound on the measured winding wheel 3g2, ensuring the winding quality.

[0154] During the inspection process, if the inspection system 7 detects any quality issues with the outer sheath of the cable 10, such as damage or cracks, the control system of the touch control system 4 will sound an audible and visual alarm and stop the rotation of the reel to be tested 3g1 and the reel to be tested 3g2, pending inspection and resolution by the operator. Once the problem has been resolved, the operator will press the start button on the control button 5 again to continue the inspection process for the cable 10.

[0155] When the speed measured by the speed measurement system 8 reaches zero, the cable 10 inspection for the reel 3g1 to be tested is concluded. The control unit in the touch control system 4 issues an audible and visual alarm, prompting the operator to press the stop button in the control button 5. The operator then replaces the reel 3g1 to be tested and the tested reel 3g2 with a new one for a new round of cable inspection. If the wound cable skin inspection device suddenly malfunctions during operation, the operator can press the emergency stop button in the control button 5 to prevent further damage to the device or the cable 10 to be tested.

Claims

1. A method for detecting the skin of a wound cable, wherein the device for detecting the skin of a wound cable comprises: A bottom bracket (1), a carrying box (2), a winding system (3), a touch control system (4), a control button (5), a wire system (6), a detection system (7), a speed measurement system (8), and a set of auxiliary brackets (9); The bottom bracket (1) is located at the bottom end of the winding cable skin detection device, and the carrying box (2) is fixed above the bottom bracket (1); on one side of the carrying box (2), a winding system (3), a wire system (6), an auxiliary bracket (9), a detection system (7), a speed measurement system (8), another auxiliary bracket (9) and another winding system (3) are installed in sequence from front to back, and the installation ends and power source parts of the wire system (6) and the two winding systems (3) are located inside the carrying box (2), and the winding ends of the wire system (6) and the two winding systems (3) are exposed outside the carrying box (2); a touch system (4) and a control button (5) are embedded and installed on the top surface of the carrying box (2); in, The touch control system (4) is used to set the working parameters of the entire device and monitor the detection status of the cable; the touch control system (4) is integrated with the control unit of the winding cable skin detection device; the touch control system (4) and the control button (5) having the control unit are located inside the carrying box (2), and the operating parts of the touch control system (4) and the control button (5) are exposed outside the carrying box (2); The control button (5) is used to start, stop and emergency stop the winding cable skin detection device; The two winding systems (3), the two wire systems (6), the detection system (7), the speed measurement system (8) and the control button (5) are simultaneously connected to a control unit of the touch control system (4), wherein the control unit is a CPU; The speed measuring system (8) is installed on the outside of the bearing box (2), between the auxiliary bracket (9) and the detection system (7), and measures the moving speed of the passing cable; the speed measuring system (8) is a speed sensor; The installation end of the winding system (3) further comprises a winding fixing seat (3a) and a winding motor bracket (3b); The power source part of the winding system (3) also includes a winding motor (3c) and a winding coupling (3d); The winding end of the winding system (3) further comprises a winding bearing seat (3e), a winding shaft (3f) and a winding wheel (3g); The winding fixed seat (3a) is a fixing and supporting component of the winding system (3) and is installed on the inner bottom surface of the bearing box (2). The winding motor bracket (3b) and the winding bearing seat (3e) are arranged and fixed on the top surface of the winding fixed seat (3a); the winding motor (3c) is fixed on the winding motor bracket (3b), the rear end of the winding shaft (3f) passes through the winding bearing seat (3e) and is connected to the output shaft of the winding motor (3c) through the winding coupling (3d); the front end of the winding shaft (3f) is installed with a winding wheel (3g); and the front end of the winding shaft (3f) passes through the side wall of the bearing box (2), and the winding wheel (3g) is located on the outside of the bearing box (2); The output shaft of the winding motor (3c) drives the winding wheel (3g) to rotate together through the winding coupling (3d) and the winding shaft (3f); The installation end of the wire system (6) further comprises a wire fixing seat (6a) and a wire motor bracket (6b). The power source part of the conductor system (6) also includes a conductor motor (6c) and a conductor coupling (6d); The winding end of the wire system (6) further comprises a wire bearing seat (6e), a wire wheel bracket (6f), a wire wheel (6g), a wire sliding block (6h), a wire screw shaft (6i), a wire support (6j), a wire sliding shaft (6k), a wire support plate (6l), and a wire lead screw nut (6m); The conductor fixing seat (6a) is a fixing and supporting component of the conductor system (6) and is installed on the inner bottom surface of the bearing box (2). The conductor motor bracket (6b) and the conductor bearing seat (6e) are arranged in sequence and fixed on the top surface of the conductor fixing seat (6a); the conductor motor (6c) is fixed on the conductor motor bracket (6b), and the rear end of the conductor screw shaft (6i) passes through the conductor bearing seat (6e) and is connected to the output shaft of the conductor motor (6c) through a conductor coupling (6d); the front end of the conductor screw shaft (6i) is provided with a screw structure, and the front end of the conductor screw shaft (6i) passes through the side wall of the bearing box (2) and extends to the outside of the bearing box (2); the conductor support plate (6l) is fixed on the outside of the bearing box (2), and two conductor supports (6j) are fixed on the conductor support plate (6l). A wire sliding shaft (6k) is installed between two wire supports (6j) at the front and rear ends of the upper surface of the plate (6l); a wire sliding block (6h) is sleeved on the wire sliding shaft (6k) and can slide back and forth along the wire sliding shaft (6k); a wire lead screw nut (6m) is embedded and installed on the wire sliding block (6h); a wire wheel bracket (6f) is connected to the top of the wire sliding block (6h); and three freely rotatable wire wheels (6g) are installed on the wire wheel bracket (6f); a portion of the wire screw shaft (6i) outside the bearing box (2) passes through the wire support (6j), the wire sliding block (6h), and another wire support (6j) in sequence, and forms a lead screw nut pair with the wire lead screw nut (6m) embedded and installed on the wire sliding block (6h); The output shaft of the conductor motor (6c) drives the conductor screw shaft (6i) to rotate together through the conductor coupling (6d); since the conductor screw shaft (6i) and the conductor lead screw nut (6m) are matched and connected in the form of a lead screw nut pair, and the conductor lead screw nut (6m) is installed and fixed on the conductor sliding block (6h), the conductor sliding block (6h) slides back and forth along the conductor sliding shaft (6k) under the action of the lead screw nut pair, thereby driving the conductor wheel bracket (6f) and the conductor wheel (6g) installed on the conductor wheel bracket (6f) to move back and forth together; The detection system (7) further comprises a detection system base (7a), a camera bracket (7b), a group of detection cameras (7c), a detection box (7d), a detection end cover (7e), and a detection system mounting frame (7f); The detection system mounting frame (7f) is fixed on the outer side of the bearing box (2), the detection system base (7a) is fixed on the detection system mounting frame (7f), and the detection box (7d) is fixed on the top of the detection system base (7a); the detection box (7d) is a hollow structure, and a group of mounting holes are distributed on the side wall of the detection box (7d), and the mounting holes are distributed in sequence at equal intervals along the circumferential direction of the cross section of the detection box (7d); each detection camera (7c) is fixed on the side wall of the detection box (7d) through a camera bracket (7b), and the shooting end of the detection camera (7c) is respectively inserted into the detection box (7d) through the mounting hole; detection end covers (7e) are installed at both ends of the detection box (7d), and the panel of the detection end cover (7e) is provided with a notch so that the detected cable can pass through; The auxiliary bracket (9) further comprises an auxiliary rod (9a) and an auxiliary wheel (9b), and the two auxiliary brackets (9) are fixed on the outer side wall of the bearing box (2) on both sides of the detection system (7); the auxiliary rod (9a) is L-shaped, and the end of one right-angled side of the auxiliary rod (9a) is fixed on the outer side wall of the bearing box (2), and the end of the other right-angled side of the auxiliary rod (9a) serves as a hanging end, and a rotating shaft is arranged on the hanging end, and a freely rotatable auxiliary wheel (9b) is installed on the rotating shaft of the auxiliary rod (9a), and the outer edge of the auxiliary wheel (9b) has a groove; The detection method comprises the following steps: Step 1: The control system controls the operation of the motors of the winding system (3) and the conductor system (6), and at the same time controls the detection camera (7c) of the detection system (7) to capture a picture of the cable surface in real time; Step 2: Use an image detection algorithm to identify and process the pictures taken at multiple angles. If damage to the epidermis is detected, the motors of the winding system (3) and the wire system (6) are controlled to stop running, and the detected damage picture is displayed to the user, who confirms whether it is line damage. After the user confirms, the control system continues to run; if not, the detection continues; Step 3: The control system simultaneously collects information from the speed sensor in real time to detect whether the current take-up speed is greater than or less than the pay-off speed. When the speed difference exceeds a preset threshold, the control system will control the corresponding motor to stop running. When the speed difference is reduced to within the preset value range, the control system will control the stopped motor to continue running. When the entire cable reel is tested, the system stops testing and controls the motor to stop. The feature is that: the process of using the image detection algorithm to identify and process the pictures taken at multiple angles in step 2 is specifically as follows: Step 21: Image preprocessing: 1.1) Read an original wire image; 1.2) Normalize the size of the wire image and convert it into a single-channel grayscale image; 1.3) Using the reduced image, the area where the wires are located is preliminarily located based on a fixed threshold; 1.4) Perform low threshold processing on the wire area in the image and high threshold processing on the background area to remove interference in the background; 1.5) Use high-pass filtering method for smoothing and mean filtering method to remove noise; Step 22: For the wire image after threshold processing, accurately search for the upper and lower edges of the wire to obtain the point sets of the upper and lower edges of the wire, thereby detecting the upper and lower edges of the wire; Step 23: Detection of wire edge anomalies; Step 24: Regularize the bent area of the wire into a rectangular area, and then detect the blocks with abnormal intensity within the area, thereby detecting the internal abnormal conditions of the wire.

2. A wound cable skin detection method according to claim 1, characterized in that: In the above-mentioned Step 22, for the wire image after threshold processing, the process of accurately searching for the upper and lower edges of the wire to obtain the point sets of the upper and lower edges of the wire, thereby detecting the upper and lower edges of the wire, includes the positioning of the left and right endpoints of the wire and the search for the wire edges. Specifically: Step 221: The process of the positioning part of the left and right endpoints of the wire is specifically as follows: First, expand the left and right boundaries of the image to provide a buffer area, assuming that the wire in the image meets the conditions of being horizontally placed and starting and ending at the left and right boundaries; After that, mask most of the middle area of the image and only extract the narrow left and right edge areas in the image; After that, perform binary processing on the left and right areas using a fixed threshold; After that, use the Shi-Tomasi corner detection method to extract the strongest 4 endpoints from the left and right areas respectively. The Shi-Tomasi corner detection includes the following steps: (a) Use the Shi-Tomasi method to calculate the change in the gray value of each pixel point in the left and right areas. The calculation method is: where E(u, v) is the gray value of the window, Ix and Iy are the partial derivatives of the gray image in the x and y directions, and w(x, y) is the window function; (b) The matrix form of E(u, v) is: Where M is the autocorrelation matrix, which is: (c) Calculate the eigenvalues λ1 and λ2 of the M matrix; (d) Determine the corner points based on the matrix eigenvalues, λ1≥λ max And λ2≥λ max , where λ max is the corner detection threshold; After that, compare the positions of the 4 endpoints in the left and right areas, and select 2 extreme points to obtain the 4 rough coordinates of the wire at the left and right boundaries of the image; Step 222: The process of the wire edge search part is specifically as follows: First, perform Canny edge detection using a low threshold to ensure that more edge information is retained (a) Perform convolution operation on the wire image using a convolution operator. * represents the convolution operation to obtain its gradient value: The magnitude G and angle θ of the gradient are calculated by the following method: (b) After obtaining the gradient magnitude and angle of the wire image, for any pixel point A in the image, if the pixel value of A is not the maximum among the pixel values in its neighborhood, then A is not an edge point of the image; if the pixel value of A is the maximum among the pixel values in its neighborhood, then A is an edge point of the image; (c) Define the high threshold H and low threshold L of the image. If the gradient G of the pixel point > H, then the pixel point is a true edge; if the gradient G of the pixel point < L, then the pixel point is a non-edge; when the gradient G of the pixel point is between L and H, judge whether there are true edge points in its surrounding neighborhood. If so, consider this point as a true edge point, otherwise consider this point as a non-edge point; After that, perform repositioning search for the left and right endpoints of the wire on the binary edge image to obtain 4 accurate coordinates; After that, from the left endpoint to the right endpoint direction, perform point-by-point accurate search for the upper and lower edges, save the searched points in an array, and use the boundary tracking method to detect the upper and lower edges of the wire; among them, the boundary tracking method adopts the following two-level logic: Level 1: Search for the next edge point within the set distance range in the order of the specified direction priority, and repeat continuously; Level 2: Determine whether the edge point searched is adjacent to the previous point. If not, there is a gap in the middle. Search for edge points point by point in the reverse priority order from the edge point searched. If there is no edge point, fill the points in the step direction until the X position of the previous point is reached. After that, after searching point by point, all the points on the upper and lower edges are finally detected. If the position of the right endpoint is not finally found, exception processing is performed.

3. A wound cable skin detection method according to claim 2, characterized in that: In the steps 2 and 3, the process of detecting abnormality at the edge of the wire is specifically as follows: Step 231: Interval sampling: calculate the height difference between adjacent points every 10 points, which represents the slope change; Assuming that point Vp1 (x1, y1) and point Vp2 (x2, y2) are separated by 10 points, calculate K_Vp1 = y2-y1, where K_Vp1 represents the slope change of point Vp1; Step 232: Calculate the height difference between the middle point and the midpoint of the line connecting the two sides for every three consecutive sampling points, which represents the concavity: Assuming that point Vp3 (x3, y3), point Vp4 (x4, y4) and point Vp5 (x5, y5) are three consecutive sampling points, M_Vp4 represents the concavity of point Vp4, M_Vp4 = (y3 + y5) / 2 - y4; Step 2, 3, 3. Smooth the slope and convexity of the wire so that there are no repeated changes or sudden abnormally high values. By setting the threshold rule, determine whether the wire has a bending problem: Set the slope change threshold KK and the concavity threshold MM. For any pixel point Vp, when K_Vp>KK, or M_Vp>MM, it is determined that the wire has a bending problem. Step 2, 3, 4: Since there may be anomalies in the left and right narrow edges of the image, shield the edge points of the left and right narrow edges of the image, calculate the average width between the upper and lower edges of the middle area, the maximum width and the minimum width of each position: Mask the edge points of the left and right narrow sides of the image. Assume that the upper and lower edge points of any position in the middle area are points Vp6 (x6, y6) and Vp7 (x7, y7). Therefore, calculate the width of the position width = y7-y6. After calculating the width of all the positions in the middle area, we get the width value set {width1, width2, ...width n }, where the maximum and minimum values ​​are denoted as max_width, min_width; Average Width Step 2, 3, 5. Compare the maximum and minimum widths with the average value to determine whether the wire width is abnormal: Normal wires do not change in width. The maximum and minimum widths are compared with the average value to determine whether the wire has width abnormalities. This method is quite effective and can effectively detect skin damage. Assuming the width face threshold WW, if |max_width-aver_width|>WW or |min_width-aver_width|>WW, it can be determined that the wire has width abnormalities.

4. A wound cable skin detection method according to claim 3, characterized in that: In the above-mentioned step 24, the area where the wire is bent is regularized into a rectangular area, and then blocks with abnormal strength are detected in the area, so as to detect abnormal conditions inside the wire. Specifically, the process is as follows: Step 241: Assume that the wires in the image are arranged in a horizontal direction. For the image, the X-axis direction is the horizontal direction and the Y-axis direction is the vertical direction. The sampling nearest neighbor interpolation method is used to normalize the vertical lines inside the wires in each X direction to the average width, and finally the wire area is extracted as a rectangular area; Step 242: Set the width value near the upper and lower edges of the shielded rectangular area to eliminate the influence of the outer edge on the inner part; Step 243: In the normal inner area of ​​the wire, the intensity changes little along the wire, so the average intensity is calculated at each Y position in the rectangular area: In the rectangular area, any point along the line is denoted as Set_j={p 1j (x1,j),p 2j (x2,j),…,p nj (x n ,j)}, where p nj (x n ,j) represents pixel p nj The coordinates of (x n ,j);I(p nj ) represents the intensity value of the point; Calculate the average intensity Calculate all j in the vertical direction within the rectangular area; Step 244: Compare each point in the rectangular area with the average intensity of its Y position. If the change exceeds the set threshold, it will be recorded as an internal abnormal point: Set the intensity threshold TT. For any point Vp(x, y) in the rectangle, when |I(Vp)-intensity_y|>TT, mark the point as an abnormal point. Step 2, 4, and 5: Agglomerate adjacent abnormal points into regions, use contour extraction to obtain the range of the abnormal region and calculate its area. When the internal abnormality exceeds a certain area, it indicates that there are patches of abnormal strength inside the wire: Set the area threshold AA, aggregate adjacent abnormal points into an area, count the number of points in the abnormal area QQ, and when QQ>AA, it is determined that there are patches with abnormal strength inside the wire.

Images