Cable parameter detection method and detection device
Through block recognition and visual recognition technology, combined with transmission transformation matrix and artificial intelligence separation algorithm, the problem of low accuracy and automation of cable parameter detection is solved, and efficient and accurate cable parameter detection is achieved, suitable for all batches on site Detection reduces costs and improves the credibility of detection data.
Patent Information
- Application Number
- CN202410861458.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-28
- Publication Date
- 2024-11-15
AI Technical Summary
The existing cable parameter detection technology has problems such as low detection accuracy, low recognition automation, low detection efficiency and low equipment accuracy. Especially in the detection of cable insulation layer thickness, it relies on manual operation and cannot achieve full on-site inspection.
Using the block recognition program and visual recognition technology, visual detection identification information is obtained through the block image, including cable type, voltage level and cross-sectional area, and using the transmission transformation matrix for accuracy correction, realizing automatic calibration and accurate measurement of insulating layers and insulation shielding. The thickness of the layer and wire core shielding layer is combined with artificial intelligence separation algorithm to realize cable type identification and parameter detection.
It improves detection accuracy and efficiency, realizes rapid detection of all batches of wires and cables on site, reduces hardware and personnel costs, enhances the credibility of inspection data, and ensures the accuracy and automation of equipment inspection.
Smart Images

Figure CN118960581A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of electric power material detection, and in particular to a cable parameter detection method and a detection device based on visual technology. Background Art
[0002] Wires and cables are one of the most important power grid materials related to economic and social development. If the insulation thickness of cables does not meet the standards, there is a risk of breakdown during use, causing power outages or even fires, resulting in significant economic losses and adverse social impacts.
[0003] Based on the random inspection results of network access in the past three years, insulation thickness and eccentricity test items accounted for 31.56% and 3.04% of all unqualified products, and the insulation thickness problem was particularly prominent.
[0004] The existing detection method detects the thickness of the insulation layer of the cable through visual technology, captures the image of the cable through a camera and a lens, and then analyzes it. For example, the existing invention patent application document "A method and system for detecting the thickness of the insulation layer of wires and cables based on machine vision" with publication number CN115014211A includes the following steps: first, configure the measurement system according to the actual measurement environment; collect and obtain the image M1 of the insulation layer and standard parts of the cable to be measured on the measurement platform through the configured measurement lens; extract the insulation layer area of the image M1 to obtain the insulation layer image M2 and preprocess it to obtain the preprocessed image M3; detect the edge of the image M3 to obtain the insulation layer edge image and perform edge refinement to obtain the image M4; use the image geometric moment algorithm to extract the center of the insulation layer in the image M4 to prepare for the insulation layer size measurement; extract the standard parts area in the image M1 and process and analyze it, and calculate the calibration coefficient K of the measurement system; finally, construct a rectangular coordinate to measure the thickness of the insulation layer of the wire and cable. And the existing invention patent application document "A cable quality inspection method based on computer vision" with publication number CN114509013A. The existing method includes the following steps: S1: cut the detected cable and use a camera to obtain a high-definition image of the cable cross section under a specified environment; S2: identify the outermost black part of the cable cross section through a computer vision algorithm and randomly locate a point a. This prior art automatically locates the center of the cable cross section through a computer vision algorithm, and automatically locates the pixel point J where the radius intersects the inner side of the insulation layer and the pixel point P where the radius intersects the outer side of the shielding layer through computer vision and edge detection algorithms, and finally automatically calculates the pixel distance between point J and point P through a coordinate conversion algorithm. The device has a fixed detection measurement unit, but as the device is used, the accuracy of the captured image will change, and the fixed detection measurement unit will be inaccurate.
[0005] The existing detection technology has the following main defects:
[0006] (1) The level of intelligent detection is low. The detection process mainly relies on manual operation. The detection results are closely related to the level of the detection personnel. For example, the insulation thickness detection requires manual use of a utility knife to cut the cable insulation layer and slice it before measurement.
[0007] (2) The testing coverage is narrow. The testing equipment involved in the testing of the insulation layer of wires and cables is large and bulky. Currently, only laboratory testing after sampling can be carried out, and it is impossible to conduct full testing at storage bases and construction sites;
[0008] (3) The credibility of test data needs to be strengthened. The recording and reporting of test data rely on manual operations, which may lead to data tampering and misjudgment.
[0009] The existing detection method uses visual technology to detect the thickness of the cable insulation layer, captures the image of the cable through a camera and lens, and then analyzes it. The equipment has a fixed detection measurement unit, but as the equipment is used, the accuracy of the collected image changes, and the fixed detection measurement unit will be inaccurate.
[0010] In summary, the existing technology has technical problems of low detection accuracy and low detection efficiency. Summary of the invention
[0011] The technical problem to be solved by the present invention is: how to solve the technical problems in the prior art of low accuracy in cable parameter detection and low degree of identification automation, low detection efficiency and low equipment precision.
[0012] The present invention adopts the following technical solution to solve the above technical problem: a cable parameter detection method comprises:
[0013] S1. Design and use a gauge block recognition program to identify the gauge block of the test sample and obtain a gauge block image;
[0014] S2. Visually identify the cross section of the test sample according to the gauge block image, obtain visual detection identification information, obtain edge information of the gauge block, obtain the imaging scale and transmission transformation matrix of the gauge block, and detect the parameters of the test sample, wherein the visual detection identification information includes: cable type, voltage level and cross-sectional area, and the parameters include: insulation shielding layer, insulation layer, and wire core shielding layer thickness;
[0015] S3. Collect the actual image of the cable cross section to detect and identify the cable cross section area, and perform regional adjustment and measurement operations according to the imaging scale of the gauge block to obtain the accuracy correction result, wherein the cable cross section area is subjected to transmission transformation according to the transmission transformation matrix to obtain the calibration imaging area; measure the pixel thickness of the insulation layer image in the calibration imaging area to correct and obtain the applicable physical thickness of the insulation layer according to the imaging scale as the accuracy correction result.
[0016] The present invention can ensure that the detection accuracy of the equipment is always accurate by identifying the standard measurement module to automatically correct the measurement accuracy of the detection equipment, increase the availability of the equipment, and reduce the cost of equipment measurement verification.
[0017] The present invention configures the size of the standard metering block in the system setting interface, develops a function of detecting and identifying the standard metering block, obtains an image after the standard metering block is placed through a camera, detects the standard metering block area with visual technology, calculates the pixel size of the upper and lower parts, compares them with the configured size of the standard metering block, calculates the imaging scale of the upper and lower parts, and automatically scales the scale when the equipment is performing detection work to ensure the accuracy of equipment detection.
[0018] The present invention distinguishes the edge of the insulation layer and the edge of the shielding layer to accurately calculate and measure the thickness of the insulation layer, the insulation shielding layer, and the wire core shielding layer. The cable parameter detection method of the present invention is applied to the intelligent detection method of the cable intelligent detector, which can enrich the intelligent detection function of the cable intelligent detector and realize the functional diversification of the cable intelligent detector.
[0019] The present invention calibrates a gauge block of known size and gives a deviation index through built-in calibration and gauge block identification. The system can automatically correct the deviation to ensure the accuracy of the detector.
[0020] In a more specific technical solution, in step S2, a model library of the inspection object is constructed, and based on the model library, the cross section of the inspection sample is visually identified to obtain the cable type, voltage level and cross-sectional area to obtain visual inspection identification information.
[0021] In a more specific technical solution, the detection objects include: wires and cables, and overhead insulated conductors.
[0022] In a more specific technical solution, in the model library, the specifications of wires, cables and overhead insulated conductors include: [0kV, 35kV].
[0023] The present invention constructs a model library of 35kV and below electric wires and cables and overhead insulated conductors, obtains parameter information such as cable type, voltage level, and cross-sectional area through visual recognition of the cross-section of the test sample, and realizes intelligent recognition of the test sample.
[0024] The thickness detection device of the present invention can replace traditional detection equipment, reduce the hardware cost and personnel cost of the detection equipment, urge manufacturers to strictly produce in accordance with national standards, and reduce the economic losses caused by production safety accidents.
[0025] In a more specific technical solution, in step S2, when the shielding layer and the insulating layer of the test sample are distinguishable, the distinguishing information is identified and the thickness of the insulating layer is calculated; otherwise, the shielding layer and the insulating layer are measured to obtain a preliminary screening basis, and the visual detection identification information is obtained based on the basis.
[0026] The present invention uses artificial intelligence visual recognition technology and adopts self-developed shielding layer, insulation layer separation and thickness measurement algorithms. It can complete cable type identification, insulation layer thickness, shielding layer thickness, eccentricity, number of cores and other parameter detection without top core, slicing and shielding layer stripping of wires and cables, shortens the detection time, improves detection accuracy and efficiency, and realizes the first on-site rapid detection of full batches of wires and cables. At the same time, the instrument is embedded with a transparent data acquisition terminal, which realizes intelligent identification of test samples, automatic judgment of test data, and remote uploading of test results, avoiding manual reporting and improving the credibility of test data.
[0027] In a more specific technical solution, step S2 further includes the following specific steps:
[0028] S21, using HSV color space conversion, convert the block image into a brightness grayscale image;
[0029] S22, performing Gaussian blur processing on the brightness grayscale image to obtain a denoised brightness grayscale image;
[0030] S23, selecting a specific gray value as a threshold, performing binarization processing on the denoised image, and obtaining a binarized image;
[0031] S24, obtaining the left and right edges of the gauge block by scanning line by line;
[0032] S25, scanning the gauge block line by line to obtain a preset number of left edge points and right edge points, and stopping the scanning when the preset number of left edge points and right edge points is reached;
[0033] S26, fitting to obtain the left edge straight line equation parameters and the right edge straight line equation parameters of the gauge block;
[0034] S27, calculating the left edge point, the right edge point and the pixel width at the upper and lower edges of the image of the block, and obtaining the imaging scale based on the calculation, making a vertical rectangular frame based on the left edge point and the right edge point, and determining the transmission transformation matrix based on the calculation;
[0035] S28. Record and calibrate the thickness of the insulating layer based on the imaging scale and the transmission transformation matrix.
[0036] In a more specific technical solution, in step S24, the method for obtaining the left and right edges of the gauge block includes: scanning the upper part of the gauge block line by line from top to bottom; scanning the lower part of the gauge block line by line from bottom to top; calculating the imaging area of the gauge block in the normal imaging position according to the preset imaging rules, and for each row of the gauge block, taking the left edge of the gauge block in the normal imaging position as the center point and a preset number of pixels as the tolerance, scanning the gauge block from left to right to determine the left edge point of the gauge block.
[0037] In a more specific technical solution, step S24 further includes:
[0038] S241, scanning the white area and the black area of the statistical block;
[0039] S242, subtract the preset value from the current scanning point position to obtain the left edge point of the current row block;
[0040] S243, execute step S241 to step S242, scan each row of gauge blocks from right to left, and determine the actual right edge of the gauge block.
[0041] In a more specific technical solution, step S241 includes:
[0042] S2411, scanning rightward to count continuous white points, and determining a white area accordingly;
[0043] S2412, continue scanning to the right to obtain the black area.
[0044] In a more specific technical solution, a cable parameter detection device includes:
[0045] A gauge block recognition module is used to design and use a gauge block recognition program to perform gauge block recognition on the test sample and obtain a gauge block image;
[0046] The gauge block visual recognition module is used to visually recognize the cross section of the test sample according to the gauge block picture, obtain visual detection recognition information, and obtain the edge information of the gauge block, thereby obtaining the imaging scale and transmission transformation matrix of the gauge block, and detecting the parameters of the test sample, wherein the visual detection recognition information includes: cable type, voltage level and cross-sectional area, and the parameters include: insulation shielding layer, insulation layer, and wire core shielding layer thickness. The gauge block visual recognition module is connected to the gauge block recognition module;
[0047] The correction result acquisition module is used to collect the actual image of the cable cross section, detect and identify the cable cross section area based on it, and perform area adjustment and measurement operations according to the imaging scale of the gauge block to obtain the accuracy correction result, wherein, according to the transmission transformation matrix, the cable cross section area is subjected to transmission transformation to obtain the calibration imaging area; the pixel thickness of the insulation layer image of the calibration imaging area is measured to correct and obtain the applicable physical thickness of the insulation layer according to the imaging scale as the accuracy correction result. The correction result acquisition module is connected to the gauge block visual recognition module.
[0048] Compared with the prior art, the present invention has the following advantages: the present invention can ensure that the equipment detection accuracy is always accurate, increase the availability of the equipment, and reduce the cost of equipment measurement calibration by identifying the standard measurement module to automatically correct the measurement accuracy of the detection equipment.
[0049] The present invention configures the size of the standard metering block in the system setting interface, develops a function of detecting and identifying the standard metering block, obtains an image after the standard metering block is placed through a camera, detects the standard metering block area with visual technology, calculates the pixel size of the upper and lower parts, compares them with the configured size of the standard metering block, calculates the imaging scale of the upper and lower parts, and automatically scales the scale when the equipment is performing detection work to ensure the accuracy of equipment detection.
[0050] The present invention distinguishes the edge of the insulation layer and the edge of the shielding layer to accurately calculate and measure the thickness of the insulation layer, the insulation shielding layer, and the wire core shielding layer. The cable parameter detection method of the present invention is applied to the intelligent detection method of the cable intelligent detector, which can enrich the intelligent detection function of the cable intelligent detector and realize the functional diversification of the cable intelligent detector.
[0051] The present invention calibrates a gauge block of known size and gives a deviation index through built-in calibration and gauge block identification. The system can automatically correct the deviation to ensure the accuracy of the detector.
[0052] The present invention constructs a model library of 35kV and below electric wires and cables and overhead insulated conductors, obtains parameter information such as cable type, voltage level, and cross-sectional area through visual recognition of the cross-section of the test sample, and realizes intelligent recognition of the test sample.
[0053] The thickness detection device of the present invention can replace traditional detection equipment, reduce the hardware cost and personnel cost of the detection equipment, urge manufacturers to strictly produce in accordance with national standards, and reduce the economic losses caused by production safety accidents.
[0054] The present invention uses artificial intelligence visual recognition technology and adopts self-developed shielding layer, insulation layer separation and thickness measurement algorithms. It can complete cable type identification, insulation layer thickness, shielding layer thickness, eccentricity, number of cores and other parameter detection without top core, slicing and shielding layer stripping of wires and cables, shortens the detection time, improves detection accuracy and efficiency, and realizes the first on-site rapid detection of full batches of wires and cables. At the same time, the instrument is embedded with a transparent data acquisition terminal, which realizes intelligent identification of test samples, automatic judgment of test data, and remote uploading of test results, avoiding manual reporting and improving the credibility of test data.
[0055] The present invention solves the technical problems existing in the prior art of low cable parameter detection accuracy, low recognition automation, low detection efficiency and low equipment precision. BRIEF DESCRIPTION OF THE DRAWINGS
[0056] Figure 1 A schematic diagram of a cable parameter detection method and detection device according to Embodiment 1 of the present invention;
[0057] Figure 2 This is a schematic diagram of a gauge block recognition interface according to Embodiment 1 of the present invention;
[0058] Figure 3 This is a schematic diagram of specific steps of visually recognizing a picture in Example 1 of the present invention;
[0059] Figure 4 This is a schematic diagram of specific steps for obtaining the left and right edges of a gauge block by scanning line by line in Example 1 of the present invention;
[0060] Figure 5 This is a schematic diagram of specific steps for counting white areas and black areas in Example 1 of the present invention;
[0061] Figure 6 This is a schematic diagram of the internal components of a detector according to Embodiment 2 of the present invention;
[0062] Figure 7 Schematic diagram of the external components of the detector according to Example 2 of the present invention. DETAILED DESCRIPTION
[0063] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in combination with the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0064] Example 1
[0065] like Figure 1 As shown, a cable parameter detection method provided by the present invention comprises the following basic steps:
[0066] S1. Design and use a gauge block recognition program to identify the gauge block of the test sample and obtain a gauge block image;
[0067] S2. Visually identify the cross section of the test sample according to the gauge block image, obtain visual detection identification information, obtain edge information of the gauge block, and obtain the imaging scale and transmission transformation matrix of the gauge block, wherein the visual detection identification information includes: cable type, voltage level and cross-sectional area;
[0068] like Figure 2 As shown, in this embodiment, a model library of wires and cables and overhead insulated conductors is constructed, and parameter information such as cable type, voltage level, and cross-sectional area are obtained through visual recognition of the cross-section of the test sample, so as to realize intelligent recognition of the test sample. In this embodiment, the specifications of wires and cables and overhead insulated conductors in the model library include but are not limited to: 35kV and below;
[0069] In this embodiment, whether to include the shielding layer in the measurement is selected according to the sample cross-section. When the cable cross-section is cut in good condition and the shielding layer and the insulation layer can be slightly distinguished, the system can identify and distinguish them and calculate the thickness of the insulation layer; otherwise, the shielding layer and the insulation layer are measured together as the basis for preliminary screening.
[0070] like Figure 3 As shown, in this embodiment, the step S2 of visually recognizing the picture also includes the following specific steps:
[0071] S21, using HSV color space conversion, convert the block image into a brightness grayscale image;
[0072] S22, performing Gaussian blur processing on the brightness grayscale image to obtain a denoised brightness grayscale image;
[0073] In this embodiment, the Gaussian filter used is a linear smoothing filter method, which blurs the image based on the characteristics of the Gaussian function. The Gaussian function is mathematically defined as:
[0074] G(x,y)=(1 / (2*π*σ 2 ))*exp(-((x-μ) 2 +(y-ν) 2 ) / (2*σ 2 ))
[0075] Where (xy) is the coordinate of the pixel, u and v are the means of the Gaussian function, and α is the standard deviation. The Gaussian function represents a two-dimensional normal distribution, and the closer the value is to the center, the greater the weight.
[0076] In the image processing of this embodiment, Gaussian filtering utilizes the central symmetry characteristic of Gaussian distribution, and achieves a smoothing effect by weighted averaging each pixel in the image with surrounding pixels, while retaining the detailed information of the image as much as possible.
[0077] In this embodiment, in step S22 of Gaussian blur processing, a smoothing operation is performed by weighted averaging, wherein for each pixel (x, y) in the image, a neighborhood window of a fixed size is taken with the pixel as the center; each pixel (i, j) in the neighborhood window is traversed, and the distance (dx, dy) between the pixel and the center pixel is calculated; according to the value of the Gaussian function, the weight coefficient w=G(dx, dy) is calculated; the pixel value in the neighborhood window is multiplied by the corresponding weight coefficient, and the sum is obtained to obtain the smoothed pixel value; the smoothed pixel value is assigned to the center pixel (x, y).
[0078] In this embodiment, the neighborhood window size is 3*3, u and v are taken as the central pixel values, σ is taken as 1.5, and Gaussian blur processing is performed on the entire brightness grayscale image to smooth the image and remove noise.
[0079] S23, selecting a specific gray value as a threshold, performing binarization processing on the denoised image, and obtaining a binarized image;
[0080] S24, obtaining the left and right edges of the gauge block by scanning line by line;
[0081] In this embodiment, in order to avoid interference at the connection of the clamp block, the contour search method is not used to obtain the accurate area of the gauge block, but the upper part is scanned line by line from top to bottom and the lower part is scanned line by line from bottom to top to obtain the left and right edges of the gauge block:
[0082] like Figure 4 As shown, in this embodiment, step S24 also includes:
[0083] S241, scanning the white area and the black area of the statistical block;
[0084] S242, subtract the preset value from the current scanning point position to obtain the left edge point of the current row block;
[0085] S243, execute step S241 to step S242, scan each row of gauge blocks from right to left, and determine the actual right edge of the gauge block.
[0086] like Figure 5 As shown, in this embodiment, step S241 includes:
[0087] S2411, scanning rightward to count continuous white points, and determining a white area accordingly;
[0088] S2412, continue scanning to the right to obtain the black area.
[0089] S25, scanning the gauge block line by line to obtain a preset number of left edge points and right edge points, and stopping the scanning when the preset number of left edge points and right edge points is reached;
[0090] In this embodiment, according to the imaging rule, the imaging area of the gauge block at the normal imaging position is calculated. For each row, the left edge of the gauge block at the normal imaging position is taken as the center point, and 50 pixels are taken as the tolerance. The actual left edge point of the gauge block is determined by scanning from left to right:
[0091] S26, fitting to obtain the left edge straight line equation parameters and the right edge straight line equation parameters of the gauge block;
[0092] In this embodiment, the initial count is 0, the count increases when a white point is scanned, and the count is cleared to 0 when a black point is scanned. When the count exceeds 3, the status is modified to that a white area has been found.
[0093] S27, calculating the left edge point, the right edge point and the pixel width at the upper and lower edges of the image of the block, and obtaining the imaging scale based on the calculation, making a vertical rectangular frame based on the left edge point and the right edge point, and determining the transmission transformation matrix based on the calculation;
[0094] In this embodiment, the initial count is 0, the count increases when a black point is scanned, and the count is cleared to 0 when a white point is scanned. When the count exceeds 3, the status is modified to that a black area has been found.
[0095] S28, recording and calibrating the thickness of the insulating layer according to the imaging scale and the transmission transformation matrix;
[0096] In this embodiment, the current scanning point position minus 4 is the left edge point of the gauge block in this row; execute the aforementioned steps S26 to S28, scan each row from right to left, and determine the actual right edge of the gauge block;
[0097] In this embodiment, scan from top to bottom line by line, and stop scanning when scanning to 5 left edge points and 5 right edge points; scan from bottom to top line by line, and stop scanning when scanning to 5 left edge points and 5 right edge points; collect 10 left edge points at the top and bottom, and fit the parameters of the equation of the straight line on the left edge of the block; collect 10 right edge points at the top and bottom, and fit the parameters of the equation of the straight line on the right edge of the block; calculate the left and right edge points and pixel widths at the top and bottom edges of the image, and calculate the imaging scale with the smaller pixel width and the actual width of the block. Make a vertical rectangular frame with the edge points at the smaller pixel width, and calculate the transmission transformation matrix based on the upper and lower 4 edge points and the rectangular frame;
[0098] In this embodiment, the vertical rectangular frame corresponds to the gauge block standard imaging rectangular frame;
[0099] In the region adjustment operation of this embodiment, the imaging scale is calculated by calculating the left and right edge points and pixel widths at the upper and lower edges of the image, and the smaller pixel width and the actual width of the gauge block are used to calculate the imaging scale. The edge points at the smaller pixel width are used to make a vertical rectangular frame, corresponding to the standard imaging rectangular frame of the gauge block, and the affine transformation matrix M is calculated based on the upper and lower 4 edge points and the rectangular frame:
[0100]
[0101] According to the affine transformation matrix, the pixel coordinate points in the cable cross-section area are affine transformed to obtain the calibration imaging area.
[0102] In this embodiment, the imaging scale and the transmission transformation matrix are recorded, and a calibration measurement of the thickness of the cable insulation layer is performed during actual measurement.
[0103] S3. Collect the actual image of the cable cross section to detect and identify the cable cross section area, and perform regional adjustment and measurement operations according to the imaging scale of the gauge block to obtain the accuracy correction result, wherein the cable cross section area is subjected to transmission transformation according to the transmission transformation matrix to obtain the calibration imaging area; measure the pixel thickness of the insulation layer image in the calibration imaging area to correct and obtain the applicable physical thickness of the insulation layer according to the imaging scale as the accuracy correction result.
[0104] In this embodiment, a transmission transformation is performed on the cable cross-sectional area according to the recorded transmission transformation matrix to obtain a calibrated imaging area; the pixel thickness of the insulation layer image in the calibrated imaging area is measured, and the accurate physical thickness of the insulation layer is calculated according to the recorded imaging scale.
[0105] In this embodiment, the detection completion system can generate a PDF detection report based on the detection results, and the detection report can be transmitted to the intranet quality management platform through WIFI or 4G communication network card for file transfer.
[0106] In the measurement operation of this embodiment, the pixel thickness of the insulating layer image in the calibration imaging area is measured, and the physical thickness of the applicable insulating layer is corrected and obtained according to the imaging scale. The corrected physical thickness of the insulating layer is:
[0107] d m =μ c ×d0
[0108] Among them, μ c is the imaging scale of the gauge block, and d0 is the pixel thickness of the insulating layer image in the measured calibration imaging area.
[0109] Example 2
[0110] like Figure 6As shown, in this embodiment, a cable parameter detection method and detection device can be used, for example, a cable intelligent detector, including:
[0111] The industrial camera 1' adopts a high-pixel area array camera to minimize the complexity of the structural design. In this embodiment, the pixel specification of the high-pixel area array camera can be, for example, greater than 12 million pixels.
[0112] The telecentric lens 2' is used to reduce the lens parallax. Within a certain object distance range, the image magnification will not change, which can effectively reduce the parallax caused by the non-vertical cable cutting plane, affecting the measurement accuracy.
[0113] LED backlight source 3', in this embodiment, the components of the light source include but are not limited to: an additional light source controller, an LED ring light source. In this embodiment, the LED ring light source is used, which can easily highlight the edge of the object to be detected at a low angle, is suitable for edge detection of the object to be detected, and improves the measurement accuracy.
[0114] The cable placement bracket 4' is installed inside the detector housing, and the adjustment knob 5' in the bracket is movably installed in the position adjustment hole of the cable placement bracket 4'; the LED front light source 6' is connected to the telecentric lens 2'.
[0115] The computing control device uses a 10-inch touch screen PAD control device, cooperates with the Android operating system, installs a WIFI module for communication, and provides a USB interface. In this embodiment, the system captures the cable cross-section, analyzes the image, measures the thickness of the N (custom) insulation layers of the cable, and calculates the average value, minimum value, and eccentricity, and makes a qualified judgment by comparing the national standard and the enterprise standard, with a detection accuracy of 0.001mm.
[0116] The control and visualization software includes a user interface for configuring the device, visualizing the detection results, and outputting the measurement results. In this embodiment, the user selects the type of cable information based on the known cable type. In this embodiment, the cable information types include but are not limited to: cable type, voltage level, and cross-sectional area. The visual algorithm is used to identify the core and calculate the number of core layers and the number of cores.
[0117] In this embodiment, the product accuracy correction system performs built-in calibration. Through gauge block recognition, the gauge blocks of known sizes are calibrated and the deviation index is given. The product accuracy correction system can automatically correct the deviation to ensure the accuracy of the detector.
[0118] In this embodiment, the product accuracy correction system has built-in national standards and enterprise standards for cable inspection, which can facilitate the review of standards and comparison of inspection results. The product accuracy correction system can determine whether it complies with national standards and enterprise standards.
[0119] The product accuracy correction system based on visual technology provided by the present invention can be applied to detection equipment including but not limited to: a detector.
[0120] like Figure 7 As shown, in this embodiment, the detector includes but is not limited to: a display screen 1, a power switch 2, a network cable interface 3, a USB interface 4, a power cable interface 5, a sliding cover 6 and a box 7.
[0121] Cable samples have a wide range of applications, including overhead cables and power cables, with inclined or partially blurred cross sections, voltage levels of 1-35kV, sample lengths of 1-15cm, and cross-sectional areas of 16-2000mm2;
[0122] In this embodiment, visual recognition can quickly identify the insulating layer and the shielding layer without the need for core top, slicing, or stripping of the shielding layer, thereby optimizing the existing inspection process and reducing the inspection time from 30-40 minutes to 3-5 minutes.
[0123] In this embodiment, the built-in analysis model can set the number of detection points according to the recognized image, and the system can accurately calculate the thickness of the insulation layer, the thickness of the shielding layer, the eccentricity, and the number of wire cores with an accuracy of 0.001mm;
[0124] In this embodiment, a quality inspection report is generated based on the measurement results, and the inspection result output is completed in 20 seconds;
[0125] In this embodiment, it is portable and adaptable to various scene inspections, and can be used in laboratories, material warehouses, and construction sites. The commissioning of the portable cable insulation thickness detector eliminates the need to manually cut 0.8-2mm insulation sheets with a utility knife and manually strip the insulation and shielding layers, which improves the safety of the inspection operators. In this embodiment, the application of the portable cable insulation thickness detector can meet the needs of on-site pre-screening, and empower the material inspection capabilities of the laboratory to the warehouse site. Due to the significant improvement in inspection efficiency and inspection capabilities, it is expected to achieve the necessary inspection of the insulation thickness of wires and cables. At the same time, it solves the problems of human resource limitations and errors that are difficult to avoid in manual operations, and has a significant effect on improving quality and efficiency. The portable cable insulation thickness detector has been demonstrated in the State Grid Anhui Material Quality Inspection Center, the State Grid Hefei Power Supply Company Material Warehouse, and the State Grid Wuhu Power Supply Company Material Quality Inspection Center. It has completed the on-site "pre-screening" of 245 reels of wires and cables across the province, and found 4 quality problems, with significant inspection results.
[0126] In summary, the present invention can ensure that the detection accuracy of the equipment is always accurate, increase the availability of the equipment, and reduce the cost of equipment measurement verification by identifying the standard measurement module to automatically correct the measurement accuracy of the detection equipment.
[0127] The present invention configures the size of the standard metering block in the system setting interface, develops a function of detecting and identifying the standard metering block, obtains an image after the standard metering block is placed through a camera, detects the standard metering block area with visual technology, calculates the pixel size of the upper and lower parts, compares them with the configured size of the standard metering block, calculates the imaging scale of the upper and lower parts, and automatically scales the scale when the equipment is performing detection work to ensure the accuracy of equipment detection.
[0128] The present invention calibrates a gauge block of known size and gives a deviation index through built-in calibration and gauge block identification. The system can automatically correct the deviation to ensure the accuracy of the detector.
[0129] The present invention constructs a model library of 35kV and below electric wires and cables and overhead insulated conductors, obtains parameter information such as cable type, voltage level, and cross-sectional area through visual recognition of the cross-section of the test sample, and realizes intelligent recognition of the test sample.
[0130] The thickness detection device of the present invention can replace traditional detection equipment, reduce the hardware cost and personnel cost of the detection equipment, urge manufacturers to strictly produce in accordance with national standards, and reduce the economic losses caused by production safety accidents.
[0131] The present invention uses artificial intelligence visual recognition technology and adopts self-developed shielding layer, insulation layer separation and thickness measurement algorithms. It can complete cable type identification, insulation layer thickness, shielding layer thickness, eccentricity, number of cores and other parameter detection without top core, slicing and shielding layer stripping of wires and cables, shortens the detection time, improves detection accuracy and efficiency, and realizes the first on-site rapid detection of full batches of wires and cables. At the same time, the instrument is embedded with a transparent data acquisition terminal, which realizes intelligent identification of test samples, automatic judgment of test data, and remote uploading of test results, avoiding manual reporting and improving the credibility of test data.
[0132] The present invention distinguishes the edge of the insulation layer and the edge of the shielding layer to accurately calculate and measure the thickness of the insulation layer, the insulation shielding layer, and the wire core shielding layer. The cable parameter detection method of the present invention is applied to the intelligent detection method of the cable intelligent detector, which can enrich the intelligent detection function of the cable intelligent detector and realize the functional diversification of the cable intelligent detector.
[0133] The present invention solves the technical problems existing in the prior art of low cable parameter detection accuracy, low recognition automation, low detection efficiency and low equipment precision.
[0134] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A cable parameter detection method, characterized in that: The method comprises: S1. Design and use a gauge block recognition program to identify the gauge block of the test sample and obtain a gauge block image; S2. According to the gauge block picture, visually identify the cross section of the test sample, obtain visual detection identification information, obtain edge information of the gauge block, obtain the imaging scale and transmission transformation matrix of the gauge block, and detect the parameters of the test sample, wherein the visual detection identification information includes: cable type, voltage level and cross-sectional area, and the parameters include: insulation shielding layer, insulation layer, and core shielding layer thickness; S3. Collect the actual image of the cable cross section to detect and identify the cable cross section area, perform area adjustment and measurement operations according to the imaging scale of the gauge block, and obtain the accuracy correction result, wherein, according to the transmission transformation matrix, perform transmission transformation on the cable cross section area to obtain the calibration imaging area; measure the pixel thickness of the insulation layer image of the calibration imaging area to correct and obtain the applicable physical thickness of the insulation layer according to the imaging scale as the accuracy correction result.
2. A cable parameter detection method according to claim 1, characterized in that: In the step S2, a model library of the inspection object is constructed, and based on the model library, the cross section of the inspection sample is visually identified to obtain the cable type, the voltage level and the cross-sectional area, so as to obtain the visual inspection identification information.
3. A cable parameter detection method according to claim 2, characterized in that: The detection objects include: electric wires and cables, and overhead insulated conductors.
4. A cable parameter detection method according to claim 3, characterized in that: In the model library, the specifications of the electric wires and cables and the overhead insulated conductors include: [0kV, 35kV].
5. A cable parameter detection method according to claim 1, characterized in that: In the step S2, when the shielding layer and the insulating layer of the test sample are distinguishable, the distinguishing information is identified and the thickness of the insulating layer is calculated; otherwise, the shielding layer and the insulating layer are measured to obtain a preliminary screening basis, and the visual inspection identification information is obtained based on the basis.
6. A cable parameter detection method according to claim 1, characterized in that: The step S2 further comprises the following specific steps: S21, converting the block image into a brightness grayscale image using HSV color space conversion; S22, performing Gaussian blur processing on the brightness grayscale image to obtain a denoised brightness grayscale image; S23, selecting a specific gray value as a threshold, and performing binarization processing on the denoised image to obtain a binarized image; S24, obtaining the left and right edges of the gauge block by scanning line by line; S25, scanning the gauge block line by line to obtain a preset number of left edge points and right edge points; stopping scanning when the preset number threshold of left edge points and right edge points is reached; S26, obtaining the left edge straight line equation parameters and the right edge straight line equation parameters of the gauge block through fitting; S27, calculating the left edge point, the right edge point and the pixel width at the upper and lower edges of the image of the block, and obtaining the imaging scale based on the calculation, and making a vertical rectangular frame based on the left edge point and the right edge point, and determining the transmission transformation matrix based on the calculation; S28. Record and calibrate the thickness of the insulating layer based on the imaging scale and the transmission transformation matrix.
7. A cable parameter detection method according to claim 6, characterized in that: In step S24, the method of obtaining the left and right edges of the gauge block includes: scanning the upper part of the gauge block line by line from top to bottom; scanning the lower part of the gauge block line by line from bottom to top; calculating the imaging area of the gauge block in the normal imaging position according to the preset imaging rule, and for each row of the gauge block, taking the left edge of the gauge block in the normal imaging position as the center point and a preset number of pixels as the tolerance, scanning the gauge block from left to right to determine the left edge point of the gauge block.
8. A cable parameter detection method according to claim 6, characterized in that: The step S24 further comprises: S241, scanning and counting the white area and the black area of the gauge block; S242, subtracting a preset value from the current scanning point position to obtain the left edge point of the gauge block in the current row; S243, executing step S241 to step S242, scanning each row of the gauge blocks from right to left to determine the actual right edge of the gauge blocks.
9. A cable parameter detection method according to claim 8, characterized in that: The step S241 includes: S2411, scanning rightward to count continuous white points, and determining the white area accordingly; S2412, continue scanning to the right to obtain the black area.
10. A cable parameter detection device, characterized in that: The device comprises: A gauge block recognition module is used to design and use a gauge block recognition program to perform gauge block recognition on the test sample and obtain a gauge block image; A gauge block visual recognition module is used to visually recognize the cross section of the test sample according to the gauge block image, obtain visual detection recognition information, and obtain edge information of the gauge block, thereby obtaining the imaging scale and transmission transformation matrix of the gauge block, and detecting the parameters of the test sample, wherein the visual detection recognition information includes: cable type, voltage level and cross-sectional area, and the parameters include: insulation shielding layer, insulation layer, and wire core shielding layer thickness, and the gauge block visual recognition module is connected to the gauge block recognition module; The correction result acquisition module is used to collect the actual image of the cable cross section, detect and identify the cable cross section area based on it, and perform area adjustment and measurement operations according to the imaging scale of the gauge block to obtain the accuracy correction result, wherein, according to the transmission transformation matrix, the cable cross section area is subjected to transmission transformation to obtain the calibration imaging area; the pixel thickness of the insulation layer image of the calibration imaging area is measured to correct and obtain the applicable physical thickness of the insulation layer according to the imaging scale as the accuracy correction result, and the correction result acquisition module is connected to the gauge block visual recognition module.
Citation Information
Cited By
Wire insulation layer sample measurement method and system
CN119374505A
High-speed wire harness assembling method and system based on intelligent algorithm detection
CN121903940A
