Automatic identification method of cable type and its cable intelligent detector and detection method

Abstract

The application discloses a cable type automatic identification method and a cable intelligent detector and detection method thereof. The automatic identification method is as follows: taking the center of the core imaging area as the center, a transverse narrow strip area subgraph is obtained, and the transverse narrow strip area subgraph is reduced to a high 1-pixel transverse narrow strip area subgraph in a gray average mode; then, a gray difference value is calculated from the left W-2 pixels to the right 2 pixels, and a corresponding gradient graph is obtained; whether the right half, the right half, the upper half and the lower half of the cable imaging area conform to the power cable characteristics is determined; if the determination result of more than or equal to two power cable characteristics is exceeded, it is determined that the cable in the image to be identified conforms to the power cable characteristics and belongs to the power cable, otherwise, it is determined to belong to the overhead cable. The application automatically identifies the cable type by comprehensively determining the results of the upper, lower, left and right four power cable characteristics, so that the possibility of misjudgment is almost nonexistent, and the technical problem that the existing cable intelligent detector cannot accurately identify the cable type is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to an automatic identification method, in particular to an automatic identification method of cable type, a cable intelligent detector using the automatic identification method, and an intelligent detection method of the cable intelligent detector. BACKGROUND

[0002] No matter power cable or overhead cable, there are their own specification requirements, the interpretation and representation method of the model specification of the cable is taken as an example of power cable, the model and variety of power cable mainly have the following aspects: 35kV and below power cable model and product representation method.

[0003] (1) The first letter of the Chinese pinyin in capital letters represents the insulation type, conductor material, inner protective layer material and structural characteristics. For example, Z represents paper (zhi); L represents aluminum (lv); Q represents lead (qian); F represents phase (fen); ZR represents flame retardant (zuran); NH represents fireproof (naihuo).

[0004] (2) The outer protective layer is represented by a number, which has two digits. No number represents no armored layer and no outer coating. The first digit represents the armor, and the second digit represents the outer coating. For example, thick steel wire armored fiber outer coating is represented as 41.

[0005] (3) The cable model is generally arranged in the order of: insulation material; conductor material; inner protective layer; outer protective layer.

[0006] (4) The cable product is represented by model, rated voltage and specification. The method is to add Arabic numerals to explain the rated voltage, core number and nominal cross-sectional area after the model. For example, VV42-10 3x50 represents copper core, polyvinyl chloride insulation, thick steel wire armored, polyvinyl chloride sheath, rated voltage 10kV, 3 cores, and nominal cross-sectional area 50mm2 of power cable. The format can be: category-[1: type, use] / [2: conductor] / [3: insulation] / [4: inner protective layer] / [5: structural characteristics] / [6: outer protective layer or derivative] / [7: make the number]6. Code and meaning of each part of power cable model:

[0007] (5) Category: ZR (flame retardant); NH (fireproof); BC (low smoke low halogen); E (low smoke halogen-free); K (control cable class); DJ (electronic computer); N (agricultural direct burial); JK (overhead cable class); B (cloth wire).

[0008] Therefore, no matter power cable or overhead cable, its specification has national standard, not defined by manufacturers at will. For example, Figure 1As shown, the left area is a schematic diagram of the port of the power cable, and the right area is a schematic diagram of the port of the overhead cable. The power cable comprises, from inside to outside, a core 1, a conductor shielding layer 2, an insulation layer 3, and an insulation shielding layer 4. The core 1 is a core area and comprises a plurality of inner cores 5 arranged in a plurality of concentric ring structures with one of the inner cores 5 as the center, one ring being sleeved on another ring. The overhead cable does not comprise the insulation shielding layer 4, and of course, there are some differences in respective size specifications.

[0009] The current cable intelligent detector has the function of calculating and detecting the thickness of the insulation layer through pictures, and although the accuracy needs to be improved, it is a great progress compared with the manual measurement method. However, the function is too single, lacks the cable automatic recognition function, and also lacks the shielding layer thickness calculation function, the core 1 area calculation function, and the inner core 5 quantity calculation function. SUMMARY

[0010] To solve the technical problem that the existing cable intelligent detector cannot accurately recognize the cable type, the present application provides an automatic recognition method of the cable type, a cable intelligent detector adopting the automatic recognition method, and an intelligent detection method of the cable intelligent detector.

[0011] The present application adopts the following technical scheme: an automatic recognition method of the cable type, comprising the following steps:

[0012] Step S10, recognizing the cable imaging area and the core imaging area in the image to be recognized;

[0013] Step S17, taking the center of the core imaging area as the center, laterally acquiring a narrow strip area with a width of W pixels and a height of h pixels to obtain a lateral narrow strip area subgraph, and reducing the lateral narrow strip area subgraph into a lateral narrow strip area subgraph with a height of 1 pixel in a gray average manner, wherein the width of the cable imaging area is W pixels, and the value of h satisfies that it is less than half of the height of the core imaging area of the smallest specification cable under the same imaging condition;

[0014] Step S18, calculating the gray difference value of the reduced lateral narrow strip area subgraph from the left W-2 pixels to the next 2 pixels to obtain a corresponding gradient graph;

[0015] Step S19, traverse all gradient values of the gradient map from left to right, respectively find a plurality of maximum value points of maximum positive change and maximum negative change, define the maximum positive change maximum value point as positive, define the maximum negative change maximum value point as negative, if the left half cable imaging area to the middle part of the core imaging area conforms to the positive-negative-positive-negative change rule, and sequentially number the corresponding maximum value points 1-4, if the distance between the numbered 3-2 points is greater than the distance between the numbered 2-1 points and the distance between the numbered 4-3 points, it is determined that the left side of the cable imaging area conforms to the power cable characteristics, and similarly traverse all gradient values of the gradient map from right to left, determine that the right half of the cable imaging area conforms to the power cable characteristics;

[0016] Step S110, referring to steps S17 to S19, longitudinally perform the acquisition of the longitudinal narrow strip area subgraph, the acquisition of the corresponding gradient map, and the analysis of the corresponding gradient value, so as to evaluate whether the upper half and the lower half of the cable imaging area conform to the power cable characteristics;

[0017] Step S111, comprehensively determine the four power cable characteristics of the four half sides, if more than or equal to two power cable characteristics are determined, it is determined that the cable in the image to be identified conforms to the power cable characteristics and belongs to the power cable, otherwise it is determined to belong to the overhead cable.

[0018] As a further improvement of the above scheme, the cable imaging area recognition method comprises the following steps:

[0019] Step S11, convert the image to be identified into a brightness grayscale image, wherein the image to be identified is a front view image obtained by imaging the cable port of the cable at a target focal length and a target imaging distance under a front view angle;

[0020] Step S13, according to the picture position and picture diameter of the cable port in the reference image, circle the corresponding picture area I on the brightness grayscale image, count the minimum value of the pixel grayscale outside the picture area I as a threshold, and then perform binaryzation processing on the brightness grayscale image; wherein the reference image is a front view image obtained by imaging the cable port of the largest size specification cable at the target focal length and the target imaging distance under a front view angle;

[0021] Step S14, use a contour finding function to find the external overall contour of the cable port in the binaryzation processed image to be identified, and then find the corresponding overall area frame, which is the cable imaging area.

[0022] Further, in step S11, the HSV color space conversion technology is used to convert the image to be identified into a brightness grayscale image.

[0023] Further, before step S13, the following step is further included:

[0024] Step S12, denoising the luminance grayscale image to remove noise points.

[0025] Further, in step S13, the minimum value of the pixel grayscale outside the picture area and located at one corner of the luminance grayscale image is counted as the threshold value.

[0026] Further, the identification method of the wire core imaging area includes the following steps:

[0027] Step S15, at the center of the cable imaging area, the n×n pixel area position is acquired, the value of n satisfies that the area of n×n is always within the area of the cable imaging area, the minimum grayscale value in the part of the pixels with the grayscale value greater than N in the rectangular area subgraph is counted as another threshold value, and the image in the cable imaging area is binarized by using the threshold value.

[0028] Step S16, the wire core area contour of the cable is found in the cable imaging area after the binarization by using a contour finding function, and the corresponding wire core area frame is calculated, that is, the wire core imaging area.

[0029] As a further improvement of the above scheme, the distance between the two points numbered 3-2 is at least 3 times the distance between the two points numbered 2-1 and the distance between the two points numbered 4-3.

[0030] As a further improvement of the above scheme, the automatic identification method further includes a cable model identification method, and the cable model identification method is:

[0031] Under the premise of identifying the cable type, the cross-sectional area of the cable is compared with the cross-sectional areas of various cables of the corresponding cable type in the cable library, and the corresponding cable model in the cable library is taken as the cable model of the cable according to the matched cross-sectional area.

[0032] The application further provides a cable intelligent detector which adopts the automatic identification method of any cable type.

[0033] The application further provides an intelligent detection method of a cable intelligent detector, which includes the automatic identification method of the cable type.

[0034] Compared with existing technologies, this invention can automatically identify cable types, eliminating the need for manual input by users and reducing input errors. Furthermore, compared to existing automatic cable type identification methods, this invention can still correctly identify cable types even if the image received is slightly obstructed or has imperfections. Moreover, this invention automatically identifies cable types by integrating the judgment results of four power cable characteristics (top, bottom, left, and right), thus virtually eliminating the possibility of misjudgment and solving the technical problem that existing intelligent cable detectors cannot accurately identify cable types. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the ports of existing power cables and overhead cables.

[0036] Figure 2 This is a flowchart of the automatic cable type identification method provided in Embodiment 1 of the present invention.

[0037] Figure 3 for Figure 2 The flowchart shows the identification method for cable imaging area and wire core imaging area used in the automatic identification method.

[0038] Figure 4 for Figure 2 A schematic diagram of image processing in the automatic recognition method.

[0039] Figure 5 This is a flowchart of a cable parameter detection method provided in Embodiment 2 of the present invention.

[0040] Figure 6 This is a flowchart of the cable core counting method provided in Embodiment 3 of the present invention.

[0041] Figure 7 for Figure 6 A schematic diagram of the equivalent rectangle II formed by the inner core counting method in the diagram.

[0042] Figure 8 This is a flowchart of the intelligent detection method of the cable intelligent detector provided in Embodiment 4 of the present invention. Detailed Implementation

[0043] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] Example 1

[0045] The embodiment discloses a cable type automatic identification method for automatically identifying whether a cable belongs to a power cable or an overhead cable, and further identifying a specific model of the cable after identifying the cable type. The cable type automatic identification method can be embedded in a CPU of a cable intelligent detector in the form of software, so that the cable intelligent detector can implement the cable type automatic identification method during detection.

[0046] Please refer to Figure 2 The cable type automatic identification method comprises the following steps.

[0047] Firstly, in step S10, a cable imaging area and a core imaging area are identified in a to-be-identified image.

[0048] Please refer to Figure 3 There are many methods for identifying the cable imaging area and the core imaging area, but a self-developed method for identifying the cable imaging area and the core imaging area is adopted in the present application, which comprises the following steps S11-S16.

[0049] In step S11, the to-be-identified image is converted into a brightness gray image.

[0050] The to-be-identified image is an image taken at a cable port, and is a front view image taken at a front view angle. An existing cable intelligent detector can be used to take an image of the cable port at a designed focal length. The focal length is a designed focal length, the distance between the camera and the cable port is a predetermined distance, and the cable port is located in the center area of the image taking picture as much as possible. The advantage of this is that a front view image of the cable port of the largest specification cable (whether it is a power cable or an overhead cable, the largest diameter cable) can be obtained at a fixed focal length and a fixed image taking distance. The parameters of the front view image are known, which provides a reference for subsequent automatic identification of the cable type.

[0051] Therefore, the to-be-identified image in step S11 is a front view image taken at a front view angle at a designed focal length and an image taking distance (i.e. at a target focal length and a target image taking distance). In actual operation, the cable intelligent detector can be provided with a positioning frame for positioning the cable and positioning the cable port in the center area of the camera picture of the cable intelligent detector, and the distance between the cable port and the camera is a target distance. Thus, the camera can take a front view image of the cable port in the center area of the picture at a front view angle at a target focal length and a target image taking distance.

[0052] The conversion of the luminance gray scale image can be performed by using the HSV color space conversion technology, and can also be performed by using image inversion, logarithmic transformation, gamma transformation, etc. as long as the luminance gray scale conversion can be performed.

[0053] In step S12, the luminance gray scale image is denoised to remove the noise points. Of course, this step can be skipped, but the denoising processing can improve the accuracy of subsequent recognition. The removal of the noise points can be performed by using the Gaussian blur calculation method. Gaussian blur, also known as Gaussian smoothing, is a processing effect widely used in image processing software such as Adobe Photoshop, GIMP, and Paint.NET. Its function is to make the image blurred and smooth, and it is usually used to reduce image noise and reduce the level of detail. The image generated by this blurring technique has a visual effect that looks like the image is observed through a frosted glass, which is obviously different from the out-of-focus imaging effect of the lens, and the effect in the ordinary lighting shadow. Gaussian filter is a linear smoothing filter, which is suitable for eliminating Gaussian noise and is widely used in the noise reduction process of image processing. In simple terms, Gaussian filtering is a process of weighted average of the entire image. The value of each pixel point is obtained by weighted average of itself and other pixel values in the neighborhood. The Gaussian blur calculation method will not be described in detail here.

[0054] In step S13, a corresponding picture area I is circled on the denoised luminance gray scale image according to the picture position and the picture diameter of the cable port in the reference image, the minimum value of the pixel gray scale outside the picture area I is counted as a threshold, and the denoised luminance gray scale image is binarized.

[0055] This reference image can be an image stored in advance in the cable intelligent detector. The reference image is a front view image obtained by imaging the cable port of the largest size specification cable at the target focal length and the target imaging distance with a front view angle. Since the focal length and the imaging distance used in the to-be-recognized image and this reference image are the same, the cable port in the center area of the imaging picture has the same size ratio and approximately the same center point. Since the cable in the reference image is the largest in size specification, the corresponding picture area I in the denoised luminance gray scale image is defined by the picture position and the picture diameter of the cable port in the reference image, and the picture area I obtained must contain the actual imaging area of the cable port of the to-be-recognized cable. Therefore, the area outside the picture area I must be all background. By using the minimum value of the pixel gray scale outside the picture area I as a threshold to perform binarization, the to-be-recognized image after binarization can distinguish the imaging of the cable port from the background, thereby facilitating the subsequent recognition of the overall outline of the cable port.

[0056] In actual operation, the minimum value of the pixel gray scale of one of the parts outside the picture area 1 can be selected as the threshold value. As shown in Figure 4 the image is taken, the left side is slightly blocked, resulting in incomplete imaging of the cable port, but one of the advantages of the present application is that it does not affect the subsequent identification of the cable type and the identification of the model specification even in such a case. For the case of Figure 4 , the minimum value of the pixel gray scale outside the cable port imaging area (i.e., outside the picture area 1) in the upper right part of the picture can be counted.

[0057] In step S14, the contour finding function is used to find the overall contour of the cable port in the binary processed image to be identified, and the corresponding overall area frame 6 is calculated, i.e., the cable imaging area.

[0058] The contour finding function is widely used, and a typical application is to define the boundary between the imaging and the background in PS image processing. The contour finding function, such as the OpenCV contour function, corresponds to a series of points, which represent a curve in the image in a certain way. In OpenCV, a contour function is represented by a series of two-dimensional vertices, and cv2.findContours() calculates the contour from a two-dimensional image. The image processed by it can be an image with edge pixels obtained from the cv2.Canny() function, or an image obtained from the cv2.threshold() and cv2.adaptiveThreshold() functions, in which case the edge is the boundary between the positive and negative areas. The overall contour finding technique will not be described in detail here, and the focus of the present application is not to design the contour finding function, but to call the existing contour finding function to find the overall contour of the cable port outside the binary processed image to be identified. After the overall contour is obtained, the points of the outermost edge in the positive direction of the overall contour can be set to set the corresponding overall area frame 6, such as the wireframe of the outermost edge in Figure 4 . The cable imaging area of the cable port is inside the overall area frame 6.

[0059] In step S15, the rectangular region subgraph is obtained at the center of the cable imaging area, and the minimum gray scale value of the part of the pixels with a gray scale value greater than N in the rectangular region subgraph is counted as another threshold value, so as to perform binary processing on the image in the cable imaging area.

[0060] The value of n satisfies that the area of n*n is always within the area of the cable imaging area, so the value of n cannot be too large, otherwise the rectangular region subgraph cannot be filled with the wire core, and in this embodiment, n=100 is sufficient. N is generally taken as 128 after evaluation according to experience. The concept of this step is similar to step S13 and will not be repeated here.

[0061] Step S16, using a contour finding function, find the contour of the core area of the cable in the cable imaging area after the binarization processing, and obtain the corresponding core area frame 7, that is, the core imaging area, as shown in the second line frame. The concept of this step is similar to step S14, which will not be repeated here. Figure 3

[0062] Step S17, taking the center of the core imaging area as the center, transversely obtaining a narrow strip area with a width of W pixels and a height of h pixels, obtaining a transverse narrow strip area subgraph, and reducing the transverse narrow strip area subgraph to a transverse narrow strip area subgraph with a height of 1 pixel by averaging the grayscale. The width of the cable imaging area is W pixels, and the value of h satisfies less than half of the height of the core imaging area of the smallest specification cable under the same imaging condition.

[0063] In this embodiment, h is 100, which is an empirical value. Here, the average refers to the average of the pixel values.

[0064] Step S18, for the reduced transverse narrow strip area subgraph, calculate the grayscale difference value of the interval of 2 pixels from the left W-2 pixels, and obtain the corresponding gradient graph.

[0065] Step S19, traverse all gradient values of the gradient graph from left to right, and respectively obtain a plurality of points with maximum positive gradient value and maximum negative gradient value. Define the point with maximum positive gradient value as positive, and define the point with maximum negative gradient value as negative. If the left half of the cable imaging area to the middle part of the core imaging area or the right half of the cable imaging area to the middle part of the core imaging area meets the positive-negative-positive-negative change rule, and the corresponding points are sequentially numbered 1-4, if the distance between the points numbered 2-1 and 3-2 is greater than the distance between the points numbered 2-1 and 3-2, and the distance between the points numbered 4-3, it is determined that the left half of the cable imaging area meets the power cable characteristic or the right half of the cable imaging area meets the power cable characteristic. In this embodiment, the standard of much greater than is at least 3 times the distance.

[0066] Step S110, referring to steps S17 to S19, longitudinally perform the acquisition of the longitudinal narrow strip area subgraph, the acquisition of the corresponding gradient graph, and the analysis of the corresponding gradient value, so as to evaluate whether the upper half and the lower half of the cable imaging area meet the power cable characteristic.

[0067] Step S111, comprehensively considering the four power cable characteristic discrimination results of the upper, lower, left and right four halves, if more than or equal to two power cable characteristic discrimination results, it is determined that the cable in the to-be-identified image meets the power cable characteristic and belongs to the power cable, otherwise it is determined to belong to the overhead cable.

[0068] ​Once the cable type is identified, the cross-sectional area of the cable is compared with the cross-sectional areas of various cables in the cable library, and the cable model corresponding to the cross-sectional area is quickly identified. In this embodiment, the cable model identification method is as follows: on the premise that the cable type is identified, the cross-sectional area of the cable is compared with the cross-sectional areas of various cables of the corresponding cable type in the cable library, and the corresponding cable model in the cable library is taken as the cable model of the cable according to the matched cross-sectional area.

[0069] The present application can automatically identify the cable type without manual input, reducing input errors. Compared with the existing cable type automatic identification method, the cable type automatic identification method of the present application can normally identify the image even if it is slightly blocked or has defects. Moreover, the present application automatically identifies the cable type by comprehensively considering the identification results of the four power cable features of the upper, lower, left, and right halves, so there is almost no possibility of misjudgment.

[0070] Through the cable type automatic identification method of the present application, the intelligent detection method of the cable intelligent detector can be applied to enrich the intelligent detection function of the cable intelligent detector and realize the functional diversification of the cable intelligent detector.

[0071] Embodiment 2

[0072] The present embodiment discloses a parameter detection method of a cable, which can include the thicknesses of the insulation shielding layer, the insulation layer, and the core shielding layer, and relates to a calculation method of the insulation layer thickness of the cable, a calculation method of the insulation shielding layer thickness of the cable, and a calculation method of the conductor shielding layer thickness of the cable. The parameters can also include the area of the core, and relate to a calculation method of the area of the core.

[0073] Please refer to Figure 5 The parameter detection method of the cable of the present embodiment includes the following steps.

[0074] Step S11, converting the to-be-identified image into a luminance grayscale image, wherein the to-be-identified image is a front view image obtained by imaging the cable port of the cable at a target focal length and a target imaging distance at a front view angle.

[0075] Step S12, denoising the luminance grayscale image to remove noise points.

[0076] Step S13, according to the picture position and picture diameter of the cable port in the reference image, a corresponding picture area one is circled on the brightness gray scale image, the minimum value of the pixel gray scale outside the picture area one is counted and taken as a threshold value, and the brightness gray scale image is binarized according to the threshold value; wherein the reference image is a front view image obtained by imaging the cable port of the cable with the largest size specification at the target focal length and the target imaging distance and under the front view angle.

[0077] Step S14, the overall contour of the cable port is found in the to-be-recognized image after the binarization processing by using a contour finding function, and a corresponding overall area frame is calculated according to the contour, that is, a cable imaging area.

[0078] Step S15, a rectangular area subgraph is obtained at the n×n pixel area position in the center of the cable imaging area, the value of n satisfies that the area of n×n is always within the area of the cable imaging area, the lowest gray scale value in the part of the pixels with the gray scale value greater than N in the rectangular area subgraph is counted as another threshold value, and the image in the cable imaging area is binarized according to the threshold value.

[0079] Step S16, the contour of the wire core area of the cable is found in the cable imaging area after the binarization processing by using a contour finding function, and a corresponding wire core area frame is calculated according to the contour, that is, a wire core imaging area.

[0080] Steps S11-S16 can be respectively referred to steps S11-S16 of embodiment 1, and will not be repeated here.

[0081] Step S27, the center points of the overall area frame 6 and the wire core area frame 7 are calculated respectively, the center point of the overall area frame 6 is moved to the center point of the wire core area frame 7, and the size of the overall area frame 6 is adjusted. The adjustment method is to increase the width and height of the overall area frame 6 respectively, and the increase amount of the width and height is the absolute value of the horizontal deviation and the absolute value of the vertical deviation between the two center points respectively.

[0082] Step S28, the width and height of the adjusted overall area frame are taken as the major axis and the minor axis of an ellipse respectively, and the ellipse is unfolded into an equivalent rectangle one again, and the pixel coordinate relationship of the adjusted overall area frame is represented as:

[0083]

[0084]

[0085]

[0086]

[0087]

[0088] In the formula, (x, y) represents the coordinates of the pixel on the adjusted overall region frame; θ is the angle of (x, y) on the elliptical coordinate system, (x1, y1) is the coordinates of the pixel on the equivalent rectangle one, a is the long axis of the ellipse, and b is the short axis of the ellipse.

[0089] In step S29, for the equivalent rectangle region one, the width and height are respectively W2 and H2, the region is divided into a narrow strip region every 50 pixels from top to bottom, a total of H2 / 50 narrow strip regions, a subgraph of each narrow strip region is intercepted, and the narrow strip region subgraph is reduced to a narrow strip region subgraph with a height of 1 pixel in a gray average manner. Wherein, W2=a, H2=2*a+2*b.

[0090] In step S210, the gray difference value of each pixel on the left side of the narrow strip region subgraph is calculated, and the corresponding gradient graph is obtained.

[0091] In step S211, all gradient values of the gradient graph of each narrow strip region subgraph are traversed from left to right, and the maximum points of multiple gradient values with the maximum positive change and the maximum negative change are obtained. If the multiple gradient maximum points from the outermost edge of the cable imaging region to the middle part of the core imaging region meet the positive (coded as No. 1), negative (coded as No. 2), positive (coded as No. 3), and negative (coded as No. 4), and each gradient maximum point is within the predetermined interval of the overall region frame 6 and the core region frame 7 (i.e. near the adjusted cable imaging region and the core imaging region), it is determined as the edge of the insulation layer and the edge of the shielding layer. In this embodiment, "nearby" means that the interval between each gradient maximum point and the nearest frame is less than one tenth of the interval between two frames. The interval between the adjacent two positive and negative gradient maximum points is determined as the edge of one layer.

[0092] If the cable is a power cable, the adjacent two positive and negative gradient maximum points are determined as the insulation shielding layer, the insulation layer, the core shielding layer, and the core from left to right. If the cable is an overhead cable, the adjacent two positive and negative gradient maximum points are determined as the insulation layer, the core shielding layer, and the core from left to right.

[0093] In step S212, the distance between the two edges of each layer is determined under the pixel coordinate relationship, and the corresponding thickness is calculated according to the distance multiplied by the proportional relationship between the pixel and the actual size.

[0094] The proportion relationship between the pixel and the actual size is calculated according to the imaging rule, and the thicknesses of the insulation layer and the shielding layer are calculated according to the found insulation layer edge and shielding layer edge. The imaging rule in this embodiment refers to the fixed focal length, the imaging distance and the proportion relationship of the actual size. According to the pixel coordinate relationship, the coordinates of the insulation layer edge and the shielding layer edge in the image to be recognized are known, so that the corresponding image thicknesses between the insulation layer, the insulation shielding layer and the conductor shielding layer in the image to be recognized are obtained, and then the actual thicknesses of the insulation layer, the insulation shielding layer and the conductor shielding layer are obtained by multiplying the proportion relationship between the pixel and the actual size.

[0095] In step S213, one layer to be calculated is defined as the A layer, the A layer thicknesses of H2 / 50 narrow strip regions are counted, the narrow strip region with the minimum A layer thickness is found out, and the data of other narrow strip regions are obtained every 60 degrees based on the narrow strip region, and a total of 6 A layer measurement data of equally spaced narrow strip regions are returned.

[0096] According to the national standard, 6 measurement data are calculated every 60 degrees. For example, the insulation layer thicknesses of H2 / 50 narrow strip regions are counted, the narrow strip region with the minimum thickness is found out, and the data of other narrow strip regions are obtained every 60 degrees based on the narrow strip region, and a total of 6 insulation layer thickness measurement data of equally spaced narrow strip regions are returned. Therefore, the final obtained insulation layer thickness measurement data has 6. By analogy, the cable intelligent detector can provide 6 groups of data, each group of data including the thicknesses of the insulation layer, the insulation shielding layer and the core shielding layer, and the measurement relationship between adjacent two groups of data is 60 degrees apart on the cable port.

[0097] The application can accurately calculate and measure the thicknesses of the insulation layer, the insulation shielding layer and the core shielding layer by distinguishing the insulation layer edge and the shielding layer edge. The parameter detection method of the cable can be applied to the intelligent detection method of the cable intelligent detector, so that the intelligent detection function of the cable intelligent detector can be enriched, and the function diversification of the cable intelligent detector can be realized.

[0098] The detection method of the cable intelligent detector can also calculate the area of the core. Specifically, the core imaging area is processed as an ellipse, and the width and height of the core imaging area are respectively taken as the major axis and the minor axis of the corresponding ellipse. The area of the ellipse corresponding to the core imaging area is calculated according to the proportion relationship between the pixel and the actual size, and is taken as the area of the core.

[0099] Embodiment 3

[0100] The embodiment discloses a core counting method of a cable, please refer to Figure 6 The core counting method of the cable of the embodiment includes the following steps.

[0101] Step S11, converting the image to be identified into a luminance gray scale image, wherein the image to be identified is a front view image obtained by imaging the cable port in the center area of the picture at a target focal length and a target imaging distance with a front view angle.

[0102] Step S12, denoising the luminance gray scale image to remove noise points.

[0103] If the cable port is not necessarily in the center area of the picture, step S13 is performed, a corresponding picture area I is circled on the luminance gray scale image according to the picture position and the picture diameter of the cable port in the reference image, the minimum value of the pixel gray scale outside the picture area I is counted as a threshold value, and the luminance gray scale image is binarized in this way; wherein the reference image is a front view image obtained by imaging the cable port of the largest size specification cable at the target focal length and the target imaging distance with a front view angle.

[0104] Step S14, using a contour finding function to find the overall contour of the cable port in the image to be identified after binarization, and to find the corresponding overall area frame, i.e. the cable imaging area.

[0105] Step S15, obtaining a rectangular region sub-image at the center n x n pixel region position of the luminance gray scale image, the value of n satisfying that the area of n x n is always within the area of the cable imaging area, counting the lowest gray scale value in the part of the pixels with a gray scale value greater than N in the rectangular region sub-image as another threshold value, and binarizing the luminance gray scale image in this way.

[0106] Step S16, using a contour finding function to find the wire core area contour in the luminance gray scale image after binarization, and to find the corresponding wire core area frame, i.e. the wire core imaging area.

[0107] Steps S11-S16 can be referred to steps S11-S16 of embodiment 1, which will not be repeated here.

[0108] Please refer to Figure 7 Step S38, taking the width and height of the wire core imaging area (i.e. the wire core area frame 7) as the major axis and the minor axis of an ellipse respectively, and unfolding the formed ellipse into an equivalent rectangle II.

[0109] Step S39, reducing the equivalent rectangle II to 1 column to the left in a gray scale average manner, counting the minimum gray scale value I therein to obtain the horizontal segmentation line of each layer of wire core in the horizontal direction.

[0110] Step S310, divide the two equivalent rectangles into subgraphs of each layer with the transverse division line as the boundary, and reduce each layer to one row in the vertical direction by the gray scale average method, and count the two gray scale minimum values.

[0111] Step S311, count the number of pixels with a gray scale minimum value less than the target threshold value as the gap between the inner cores of the wire core.

[0112] Step S312, calculate the distance between the two gray scale minimum values, and count the number of the same distance by histogram, and take the maximum number of distance values as the width of the inner core.

[0113] Step S313, calculate the number of inner cores of each layer according to the total width of the two equivalent rectangles minus the gap number of the inner core, and then divided by the width of the inner core.

[0114] Step S314, aggregate the multiple layers to obtain the total number of inner cores.

[0115] Please refer to Figure 7 The inner core counting method of the cable of the present application can accurately count even if the outline between the inner cores is not clear. By applying the inner core counting method of the cable of the present application to the intelligent detection method of the cable intelligent detector, the intelligent detection function of the cable intelligent detector can be enriched, and the diversification of the function of the cable intelligent detector can be realized.

[0116] Embodiment 4

[0117] The cable intelligent detector of the present embodiment can adopt the automatic recognition method of the cable type of embodiment 1, and can also adopt the parameter detection method of the cable of embodiment 2, and can also adopt the inner core counting method of the cable of embodiment 3. Please refer to Figure 8 In addition to the intelligent detection method of the cable intelligent detector, the following steps can be included.

[0118] Step S315, the cable intelligent detector can be networked through a 4G module, linked to the background, and remotely upgraded. Step S315 can ensure that the version of the cable intelligent detector is updated in time.

[0119] Step S316, the pictures that cannot be recognized by the cable intelligent detector can be returned through the 4G network, and the background can update the detector database by training. Step S316 can gradually improve the database and improve the recognition rate of the cable intelligent detector.

[0120] In the present embodiment, the detection box of the cable intelligent detector can also be designed as an integrated detector, which can reasonably utilize the space of the detection box, reduce the overall volume of the detector, and reduce the weight of the detector.

[0121] From the above four embodiments, it can be seen that the cable intelligent detector can intelligently identify the cable type, making detection more convenient and accurate. The calculation functions of the shielding layer, the core area (i.e. the wire core area), and the number of cores (i.e. the number of inner cores) are added. These parameters are important detection indicators in cable quality detection. Thus, the cable intelligent detector is more complete in function and better in performance. The 4G communication function is added, so that the cable intelligent detector can be online at all times, update the latest detection software in time, process the cable that cannot be identified in time, and improve the product detection efficiency and applicability. The lightweight design of the cable intelligent detector reduces the weight of the equipment and facilitates the on-site detection.

[0122] The advantages of the present application are:

[0123] (1) The present application provides an automatic identification method of cable type: judging whether the cable is a power cable or an overhead cable according to the different layered types of power cables and overhead cables; and then identifying the cross-sectional area of the cable, and quickly identifying the cable model according to the area comparison with the cable area of the cable library.

[0124] (2) The present application provides a parameter detection method of the cable, specifically a calculation method of the thickness of the insulating layer of the cable, a calculation method of the thickness of the insulating shielding layer of the cable, and a calculation method of the thickness of the conductor shielding layer of the cable. The thicknesses of the insulating layer and the shielding layer are accurately calculated and measured by accurately distinguishing the edges of the insulating layer and the shielding layer.

[0125] (3) The present application provides a calculation method of the area of the wire core, which takes the wire core imaging area as an ellipse, and takes the outer frame of the wire core area as the major and minor axes of the ellipse. According to the proportional relationship between the pixel and the actual size, the area of the wire core is calculated.

[0126] (4) The present application provides a method for counting the inner cores of the cable. The number of wire cores is counted according to the area by distinguishing the edges of the inner cores.

[0127] (5) The present application updates the intelligent detection method of the cable intelligent detector in real time: the equipment is connected to the background through the 4G module, and the machine can be remotely upgraded to ensure that the equipment version is the latest.

[0128] (6) The present application updates the database of the cable type or the cable model that cannot be identified by the cable intelligent detector in time. Through the 4G network, the unrecognizable pictures are returned back. The background updates the detector database by training, gradually improves the database, and improves the recognition rate of the detector.

[0129] (7) The present application uses a detection box as a box body to design an integrated detector, reasonably utilizes the space of the detection box, reduces the overall volume of the detector, and reduces the weight of the detector.

[0130] Therefore, the cable intelligent detector has the following functions: (1) a function of automatically identifying the cable type; (2) a function of identifying the thickness of the cable shielding layer; (3) a function of calculating the cross section of the core (i.e. the wire core); (4) a function of calculating the number of the core wire core (i.e. the core inside the wire core); (5) a function of remotely upgrading the cable intelligent detector by the system (i.e. the background); (6) a function of manually identifying the unrecognized picture returned by the cable intelligent detector by the system and remotely updating the database of the cable intelligent detector.

[0131] The above-mentioned embodiments only express several embodiments of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are all within the scope of protection of the present application. Therefore, the scope of protection of the present application patent should be subject to the appended claims.

Claims

1. A method of automatic recognition of a cable type, characterized by, It comprises the following steps: Step S10, identifying the cable imaging area and the core imaging area in the image to be identified; Step S17, laterally acquiring a width of W pixels, and a height of h pixels, to obtain a lateral narrow strip region subgraph, and reducing the lateral narrow strip region subgraph into a lateral narrow strip region subgraph with a height of 1 pixel in a gray scale average manner, wherein the width of the cable imaging region is W pixels, h the value of which satisfies less than half of the height of the core imaging region of the minimum specification cable under the same imaging condition. Step S18, for the reduced horizontal narrow strip region subgraph, from left W -2 pixel calculation back interval 2 pixels of gray scale difference value, get the corresponding gradient map; Step S19, traversing all gradient values of the gradient map from left to right, respectively obtaining a plurality of maximum value points of maximum positive change and maximum negative change, defining the maximum positive change maximum value point as positive, defining the maximum negative change maximum value point as negative, if the middle part of the left half cable imaging area to the core imaging area conforms to the positive-negative-positive-negative change rule, and sequentially numbering the corresponding maximum value points 1-4, if the distance between the numbered 3-2 two points is greater than the distance between the numbered 2-1 two points and the distance between the numbered 4-3 two points, it is determined that the left side of the cable imaging area conforms to the power cable characteristics, and similarly traversing all gradient values of the gradient map from right to left, it is determined that the right half of the cable imaging area conforms to the power cable characteristics; Step S110, referring to steps S17 to S19, longitudinally acquiring the longitudinal narrow strip area subgraph, acquiring the corresponding gradient map, and analyzing the corresponding gradient value, so as to evaluate whether the upper half and the lower half of the cable imaging area conform to the power cable characteristics; Step S111, comprehensively determining the four power cable characteristics of the four half sides, if more than or equal to two power cable characteristics are determined, it is determined that the cable in the image to be identified conforms to the power cable characteristics and belongs to the power cable, otherwise it is determined to belong to the overhead cable; The identification method of the cable imaging area comprises the following steps: Step S11, converting the image to be identified into a brightness grayscale image, wherein the image to be identified is a front view image obtained by imaging the cable port of the cable at a target focal length and a target imaging distance under a front view angle; Step S13, according to the picture position and picture diameter of the cable port in the reference image, the corresponding picture area one is circled on the brightness grayscale image, the minimum value of the pixel grayscale outside the picture area is counted as a threshold, and the brightness grayscale image is binarized; wherein the reference image is a front view image obtained by imaging the cable port of the largest size specification cable at the target focal length and the target imaging distance under the front view angle; Step S14, using a contour finding function to find the external overall contour of the cable port in the image to be identified after binarization, and the corresponding overall area frame is calculated, that is, the cable imaging area; The identification method of the core imaging area comprises the following steps: Step S15, obtaining a rectangular region subgraph at the pixel region position of n × n Step S16, obtaining the minimum gray value of the pixels in the rectangular region subgraph as another threshold value, and performing binaryzation processing on the image in the cable imaging region according to the threshold value. n n × n Step S17, obtaining the minimum gray value of the pixels in the rectangular region subgraph as another threshold value, and performing binaryzation processing on the image in the cable imaging region according to the threshold value. N Step S18, obtaining the minimum gray value of the pixels in the rectangular region subgraph as another threshold value, and performing binaryzation processing on the image in the cable imaging region according to the threshold value.​ Step S16, using a contour finding function to find the core area contour of the cable in the cable imaging area after binarization, and the corresponding core area frame is calculated, that is, the core imaging area.

2. The method of automatic recognition of the cable type according to claim 1, characterized in that, In step S11, the HSV color space conversion technology is used to convert the image to be identified into a brightness grayscale image.

3. The method of automatic recognition of the cable type according to claim 1, characterized in that, Before step S13, the following steps are further included: Step S12, denoising the brightness grayscale image to remove noise points.

4. The method of automatic recognition of the cable type according to claim 2, characterized in that, In step S13, the minimum value of the pixel grayscale outside the picture area and located in one corner of the brightness grayscale image is counted as the threshold.

5. The method of automatic recognition of the cable type according to claim 1, characterized in that, The distance between the two points numbered 3-2 is at least 3 times or more than the distance between the two points numbered 2-1 and the distance between the two points numbered 4-3.

6. The method of automatic recognition of the cable type according to claim 1, characterized in that, The automatic identification method further comprises a cable model identification method, wherein the cable model identification method is: Under the premise of identifying the cable type, the cross-sectional area of the cable is compared with the cross-sectional areas of various cables of the corresponding cable type in the cable library, and the corresponding cable model in the cable library is taken as the cable model of the cable according to the matched cross-sectional area.

7. A cable intelligence detector, characterized by, The cable type automatic identification method comprises the following steps:

8. A method for intelligent detection of a cable intelligent detector, characterized in that, The cable type automatic identification method comprises the following steps:

Images