A non-contact automatic monitoring method and system for elevator wire rope burrs

Abstract

The present invention discloses a non-contact automatic monitoring method and system for elevator wire rope burrs, relating to the field of monitoring. The method comprises: acquiring an image of an elevator wire rope; selecting a wire rope region of interest in the image; performing morphological processing on the wire rope region of interest to obtain a morphologically processed wire rope region of interest; performing adaptive enhancement processing on the morphologically processed wire rope region of interest to obtain a self-adaptively enhanced wire rope region of interest; performing binarization processing on the self-adaptively enhanced wire rope region of interest to obtain a binarized wire rope region of interest; performing sub-pixel edge detection on the binarized wire rope region of interest to determine a target area and a background area; and detecting the position and direction of burrs on the wire rope in the background area. The present invention reduces the influence of the operator's technical level on the detection results, can realize automatic detection, shortens detection time, and improves work efficiency.

Description

Technical Field

[0001] The present invention relates to the field of monitoring, and in particular to a non-contact automatic monitoring method and system for elevator wire rope burrs. Background Art

[0002] Elevators have become an indispensable piece of specialized equipment in people's daily lives, and the elevator's wire rope is a crucial component. The wire rope is a crucial suspension mechanism, bearing the full weight of the car and counterweight. The friction between the traction unit and the wire rope drives the elevator. During operation, the wire rope generates tensile stress, contact stress, and bending stress. Over time, wire rope defects such as wire breakage and corrosion can occur, even damaging the traction unit's rope grooves. This can alter traction conditions and affect normal elevator operation. In severe cases, wire rope breakage can occur, which, if not detected promptly, can have serious consequences for production safety. The presence of burrs on the wire rope is a key indicator of its safety. Traditional methods for inspecting elevator wire rope burrs rely primarily on visual inspection, requiring operators to inspect the wire rope for burrs on-site. This can lead to missed detections and false detections. Summary of the Invention

[0003] The purpose of the present invention is to provide a non-contact automatic monitoring method and system for elevator wire rope burrs, which can automatically process images of elevator wire ropes to obtain the burr conditions of the elevator wire ropes and improve the accuracy of burr monitoring.

[0004] To achieve the above-mentioned purpose, the present invention provides the following scheme: a non-contact automatic monitoring method for elevator wire rope burrs, the non-contact automatic monitoring method for elevator wire rope burrs comprising: acquiring an image of the elevator wire rope; selecting a wire rope region of interest in the image; performing morphological processing on the wire rope region of interest to obtain a morphologically processed wire rope region of interest; performing adaptive enhancement processing on the morphologically processed wire rope region of interest to obtain a self-adaptively enhanced wire rope region of interest; performing binarization processing on the self-adaptively enhanced wire rope region of interest to obtain a binarized wire rope region of interest; performing sub-pixel edge detection on the binarized wire rope region of interest to determine a target area and a background area; the target area is the wire rope area, and the background area is the burr area; detecting the position and direction of the burrs on the wire rope in the background area.

[0005] A non-contact automatic monitoring system for elevator wire rope burrs, the non-contact automatic monitoring system for elevator wire rope burrs comprising: an acquisition module for acquiring an image of the elevator wire rope; a selection module for selecting a wire rope region of interest in the image; a processing module for performing morphological processing on the wire rope region of interest to obtain a morphologically processed wire rope region of interest; a morphological processing model for performing adaptive enhancement processing on the morphologically processed wire rope region of interest to obtain a self-adaptively enhanced wire rope region of interest; an adaptive enhancement module for performing binarization processing on the self-adaptively enhanced wire rope region of interest to obtain a binarized wire rope region of interest; a sub-pixel edge detection module for performing sub-pixel edge detection on the binarized wire rope region of interest to determine a target area and a background area; the target area is the wire rope area, and the background area is the burr area; and a burr detection module for detecting the position and direction of the wire rope burrs in the background area.

[0006] A computer-readable storage medium stores a computer program, which, when executed, implements the non-contact automatic monitoring method for elevator wire rope burrs as described above.

[0007] According to a specific embodiment provided by the present invention, the present invention discloses the following technical effects: the present invention discloses a non-contact automatic monitoring method and system for elevator wire rope burrs; the method comprises: obtaining an image of the elevator wire rope by contactlessly obtaining the image to avoid affecting the normal operation of the elevator; selecting a wire rope region of interest in the image; performing morphological processing on the wire rope region of interest to obtain a morphologically processed wire rope region of interest; performing adaptive enhancement processing on the morphologically processed wire rope region of interest to obtain a wire rope region of interest after adaptive enhancement processing; performing binarization processing on the wire rope region of interest after adaptive enhancement processing to obtain a wire rope region of interest after binarization processing; performing sub-pixel edge detection on the wire rope region of interest after binarization processing to determine the target area and the background area; and detecting the position and direction of the wire rope burrs in the background area. The present invention can obtain the position and direction of the wire rope burrs by performing a series of processing on the non-contact obtained image, thereby solving the problems of on-site detection difficulties and missed detection in existing elevator wire rope burr detection methods, reducing the influence of the operator's technical level on the detection results, realizing automatic detection, shortening detection time, and improving work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0009] Figure 1 Flowchart of the non-contact automatic monitoring method for elevator wire rope burrs according to an embodiment of the present invention.

[0010] Figure 2 3D diagram of a non-contact automatic monitoring device for elevator wire rope burrs according to an embodiment of the present invention.

[0011] Figure 3 2. It is a front view of a non-contact automatic monitoring device for elevator wire rope burrs according to an embodiment of the present invention.

[0012] Figure 4 4 is a global two-dimensional histogram of the non-contact automatic monitoring method for elevator wire rope burrs in an embodiment of the present invention.

[0013] Figure 5 2 is a detection principle diagram of a non-contact automatic monitoring method for elevator wire rope burrs in an embodiment of the present invention.

[0014] Figure 6 Schematic diagram of edge lines of a non-contact automatic monitoring method for elevator wire rope burrs according to an embodiment of the present invention.

[0015] Explanation of symbols: 1. L-shaped background board; 2. Elevator wire rope; 3. High-speed camera; 4. Retractable background board bracket. DETAILED DESCRIPTION

[0016] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only 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 making creative efforts are within the scope of protection of the present invention.

[0017] The purpose of the present invention is to provide a non-contact automatic monitoring method and system for elevator wire rope burrs, which can automatically process images of elevator wire ropes to obtain the burr conditions of the elevator wire ropes and improve the accuracy of burr monitoring.

[0018] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] Example 1: Figure 1 As shown, the present invention provides a non-contact automatic monitoring method for elevator wire rope burrs, and the non-contact automatic monitoring method for elevator wire rope burrs includes the following steps.

[0020] Step 101: Acquire an image of the elevator wire rope, specifically comprising: installing a test device near the tested wire rope so as to capture clear local continuous images of the wire rope without affecting the operation of the elevator, and starting a high-speed camera to acquire an image of the elevator wire rope.

[0021] Step 102: Select a wire rope region of interest in the image.

[0022] Step 103: performing morphological processing on the steel wire rope region of interest to obtain the morphologically processed steel wire rope region of interest.

[0023] Step 104: performing adaptive enhancement processing on the morphologically processed wire rope region of interest to obtain the adaptively enhanced wire rope region of interest.

[0024] Step 105: Binarization is performed on the wire rope region of interest after the adaptive enhancement process to obtain a binarized wire rope region of interest. By combining image morphology and binarization, the contrast of the wire rope image is improved.

[0025] like Figure 2 and Figure 3 As shown, in a specific implementation, first install an L-shaped backlight panel 1 at the position where the wire rope is to be measured. Place the L-shaped backlight panel 1 on the ground through a retractable background panel bracket 4, and adjust the height of the retractable background panel bracket 4 so that the L-shaped backlight panel 1 does not affect the operation of the elevator wire rope 2.

[0026] Two high-speed cameras 3 are set up so that the main optical axes of the high-speed cameras 3 are approximately perpendicular. The parameters of the high-speed cameras 3 are adjusted to ensure that clear images of the entire length of the elevator wire rope 2 can be obtained simultaneously.

[0027] Select the region of interest (ROI) of the image, perform filtering and morphological preprocessing operations such as opening and closing operations on the ROI image to reduce the impact of image noise.

[0028] Adaptively enhance the image to improve the recognition of wire ropes. The specific processing method is as follows.

[0029] 1) Assume there is an M×N grayscale image with L grayscale levels. Then the image size of its neighborhood mean is also M×N and the grayscale level is also L. Then, the grayscale value of pixel i and its neighborhood mean j form a two-dimensional vector (i, j). f(i, j) represents the number of times the two-dimensional vector (i, j) appears in the global two-dimensional histogram, 0≤i≤L-1, 0≤j≤L-1. Then its joint probability density is for: .

[0030] And there are: , .

[0031] 2) Assuming that s and t are used to mark the thresholds of grayscale f(x, y) and neighborhood grayscale mean g(x, y) respectively, the threshold pair can divide the two-dimensional histogram of the image into four regions, such as Figure 4 shown.

[0032] 3) According to the definition of a 2D histogram, the first region A and the third region C on the diagonal of the 2D histogram are often regarded as the target and background, respectively, while the second region B and the fourth region D farther from the diagonal correspond to noise and edges, respectively. Assume that the global 2D histogram is divided into four regions, where: , .

[0033] in, represents the percentage of the first region divided by the first threshold s and the second threshold t in the global two-dimensional histogram; It represents the percentage of the third region divided by the first threshold s and the second threshold t in the global two-dimensional histogram.

[0034] 4) The mean vectors of the first and third regions are as follows.

[0035] .

[0036] .

[0037] in, represents the mean vector of the first region divided by the first threshold s and the second threshold t; represents the mean grayscale value of all pixels in the first region divided by the first threshold s and the second threshold t; represents the mean of the neighborhood means of all pixels in the first region divided by the first threshold s and the second threshold t; represents the mean vector of the third region divided by the first threshold s and the second threshold t; represents the mean grayscale value of all pixels in the third area divided by the first threshold s and the second threshold t; represents the mean of the neighborhood means of all pixels in the third region divided by the first threshold s and the second threshold t; T represents transpose.

[0038] 5) The mean vector of the global two-dimensional histogram is.

[0039] .

[0040] in, represents the mean vector of the global two-dimensional histogram; Represents the mean of the grayscale values of all pixels in the global two-dimensional histogram; Represents the mean of the neighborhood means of all pixels in the global 2D histogram.

[0041] 6) The probability of the second and fourth regions in the image two-dimensional histogram region partition diagram is relatively small, so the second and fourth regions are ignored in the conventional two-dimensional histogram Otsu method, that is: , .

[0042] 7) The dispersion measures for the first and third regions are defined as .

[0043] .

[0044] 8) S B The threshold s corresponding to the maximum value of (s, t) is the first optimal segmentation threshold s', and t is the second optimal segmentation threshold t', that is: .

[0045] 9) The traditional Otsu method for 2D histograms ignores the second and fourth regions, resulting in image edge distortion. The following describes an adaptive image enhancement algorithm based on the Otsu method.

[0046] 10) The adaptive image enhancement algorithm also uses the above-mentioned two-dimensional histogram construction method. The difference is that the adaptive image enhancement algorithm not only considers the first and third regions, but also the second and fourth regions. In order to better enhance the edges of the concentric circle image, instead of changing the grayscale value of the pixel to 0 or 1 as in image segmentation, the sigmoid function is used here to enhance the grayscale values of the pixels in the first and fourth regions respectively, while reducing the grayscale value of the pixels in the third region. The Sigmoid function is also called the S function, or the logistic function. It is a continuous, smooth, and strictly monotonic threshold function. Its mathematical model is: .

[0047] in, Represents the Sigmoid function.

[0048] 11) According to the properties of the Sigmoid function, The range of is (0, 1), and it has very good symmetry. Figure 4 As shown in Figure 3, the main steps of the adaptive image enhancement method are as follows.

[0049] Step 1. According to the traditional global two-dimensional histogram Otsu method, the first optimal segmentation threshold s' and the second optimal segmentation threshold t' in two dimensions of the image grayscale f(x, y) and the neighborhood grayscale mean g(x, y) are obtained.

[0050] Step 2. Divide the global two-dimensional histogram of the image pixel grayscale values into four regions using the first and second optimal thresholds. Calculate the ranges (v1, v2) of the pixel grayscale values in the first and third regions. Save the pixel grayscale values and pixel coordinates for these two regions.

[0051] Step 3. Due to the uncertainty of the grayscale values of noise pixels and the blurred edges of defocused images, the pixel values of the second and fourth regions interpenetrate each other, making it impossible to effectively distinguish between image noise and image edge pixels. Therefore, in the present invention, the maximum inter-class variance method is used to calculate the global thresholds of the second and fourth regions. Based on the global threshold of the second region, the second region is divided into a noise region and a non-noise region; based on the global threshold of the fourth region, the fourth region is divided into an edge region and a non-edge region.

[0052] Step 4. Perform the next step of processing on the noise and edge regions obtained in Step 3. First, replace the grayscale values of the pixels in the noise region with the grayscale median of all pixels within a fixed window to eliminate the noise in the image. Then, calculate the range (v3, v4) of the grayscale values of the pixels in the edge region. Finally, use the Sigmoid function to map (v3, v4) to (v1, v2), as shown below: , .

[0053] Most image segmentation and enhancement algorithms simply segment image pixel grayscale values into 0 or 1, or manually select an enhancement threshold. This method automatically selects a threshold and maps the grayscale values of edge regions to the target region of the global image using a Sigmoid function that better matches the image's edge structure. Compared to other methods, this method is simple to implement and offers high image restoration accuracy.

[0054] Step 106: Sub-pixel edge detection is performed on the binarized wire rope region of interest to determine the target region and the background region. The target region is the wire rope region, and the background region is the burr region. The sub-pixel edge detection algorithm accurately delineates the wire rope and burr regions.

[0055] Step 107: Detect the position and direction of the burrs on the wire rope in the background area. The number of burrs on the wire rope is detected through accurate sub-pixel edge detection.

[0056] In practice, pixel elements (CCD, CMOS) are light integrators. Their main principle is to integrate the light intensity projected onto a pixel of a fixed size within a fixed time interval. The output is the sampled value of the light intensity. Ideally, the edge projection of an object will split the sub-pixel unit where it is located into two parts with different light intensities, such as Figure 5 Therefore, a reasonable assumption is that the light intensity of the pixel matrix in the background area where the edge line passes through conforms to the area weighted average principle, which can be expressed as .

[0057] in, Represents the grayscale value of the pixel in the i'th row and j'th column in the background area pixel matrix; is the area corresponding to the part with gray value C in the background area pixel matrix, is the area corresponding to the part with gray value 0 in the background area pixel matrix, h is the side length of the background area pixel matrix, .

[0058] Further sorting can be obtained: .

[0059] like Figure 2 and Figure 3 As shown, to improve the accuracy of detection results, a test device must be designed, manufactured, and installed. The test device includes: an L-shaped background plate 1, a retractable background plate bracket 4, and at least two high-speed cameras 3. The L-shaped background plate 1 is connected to the retractable background plate bracket 4, and the two high-speed cameras 3 are arranged vertically. The L-shaped background plate 1 is installed in a suitable position to not affect the operation of the elevator wire rope 2 while capturing clear images of the wire rope. The high-speed cameras 3 can capture images of the entire wire rope, improving detection results and efficiency.

[0060] Before acquisition, the proportional factor between the actual diameter length of the wire rope and the pixel width is calculated based on the known wire rope diameter. The actual distance is calculated using the proportional factor method, so that the image can be acquired more clearly.

[0061] The present invention includes burr detection of one steel wire rope but is not limited to one steel wire rope. Multiple steel wire ropes can be detected simultaneously by selecting multiple ROIs. The present invention also includes a non-contact automatic detection method for elevator steel wire rope burrs but is not limited to a non-contact automatic detection method for elevator steel wire rope burrs.

[0062] As a specific embodiment, performing adaptive enhancement processing on the wire rope region of interest after morphological processing to obtain the wire rope region of interest after adaptive enhancement processing specifically includes the following steps.

[0063] A global two-dimensional histogram of the morphologically processed wire rope region of interest is constructed; the element f(i, j) in the global two-dimensional histogram represents the number of occurrences of a two-dimensional vector (i, j), and the two-dimensional vector (i, j) is composed of the grayscale value i of the pixel and the neighborhood mean j of the pixel.

[0064] Based on the global two-dimensional histogram, the Otsu method is used to obtain the first optimal threshold and the second optimal threshold; the first optimal threshold is the optimal threshold of the wire rope interest region after morphological processing in the grayscale dimension, and the second optimal threshold is the optimal threshold of the wire rope interest region after morphological processing in the neighborhood grayscale mean dimension.

[0065] The global two-dimensional histogram is divided into a first region, a second region, a third region and a fourth region based on the first optimal threshold and the second optimal threshold; the first region is a region whose horizontal coordinate is in the interval (0, s') and the vertical coordinate is in the interval (0, t'), the second region is a region whose horizontal coordinate is in the interval (0, s') and the vertical coordinate is in the interval (t', L-1), the third region is a region whose horizontal coordinate is in the interval (s', L-1) and the vertical coordinate is in the interval (t', L-1), and the fourth region is a region whose horizontal coordinate is in the interval (s', L-1) and the vertical coordinate is in the interval (0, t'); s' represents the first optimal threshold; t' represents the second optimal threshold; L represents the number of gray levels.

[0066] The maximum inter-class variance method is used to calculate the global thresholds of the second and fourth regions.

[0067] The second area is divided into a noise area and a non-noise area based on a global threshold of the second area; the noise area is an area whose horizontal coordinate is in the interval (0, s') and the vertical coordinate is in the interval (t', t1), and the non-noise area is an area whose horizontal coordinate is in the interval (0, s') and the vertical coordinate is in the interval (t1, L-1); t1 represents the global threshold of the second area.

[0068] The fourth area is divided into an edge area and a non-edge area based on a global threshold of the fourth area; the edge area is an area whose horizontal coordinate is in the interval (s', s1) and whose vertical coordinate is in the interval (0, t'); the non-noise area is an area whose horizontal coordinate is in the interval (s1, L-1) and whose vertical coordinate is in the interval (0, t'); s1 represents the global threshold of the fourth area.

[0069] Performing median processing on the grayscale values of pixels in the noise area of the second area to obtain a second area after median processing.

[0070] Based on the grayscale value range of the first region and the grayscale value range of the edge region, mapping processing is performed on the grayscale values of the pixels in the edge region of the fourth region to obtain a mapped fourth region.

[0071] The first region, the second region after the median processing, the third region and the fourth region after the mapping processing are used as the wire rope interest region after the adaptive enhancement processing.

[0072] As a specific embodiment, based on the global two-dimensional histogram, the Otsu method is used to obtain the first optimal threshold and the second optimal threshold, which specifically includes the following steps.

[0073] Based on the global two-dimensional histogram, the discrete measure function is obtained by using the Otsu method.

[0074] The first threshold and the second threshold when the discrete measure function reaches its maximum value are taken as the first optimal threshold and the second optimal threshold.

[0075] As a specific embodiment, the discrete measurement function is.

[0076] .

[0077] in, represents a discrete measure function; represents the mean grayscale value of all pixels in the first region divided by the first threshold s and the second threshold t; represents the mean of the neighborhood means of all pixels in the first region divided by the first threshold s and the second threshold t; represents the mean grayscale value of all pixels in the third area divided by the first threshold s and the second threshold t; represents the mean of the neighborhood means of all pixels in the third region divided by the first threshold s and the second threshold t; Represents the mean of the grayscale values of all pixels in the global two-dimensional histogram; represents the mean of the neighborhood means of all pixels in the global two-dimensional histogram; represents the percentage of the first region divided by the first threshold s and the second threshold t in the global two-dimensional histogram; It represents the percentage of the third region divided by the first threshold s and the second threshold t in the global two-dimensional histogram.

[0078] Before applying sub-pixel edge detection, single edge pixels in an image are calculated using traditional derivative masks, such as partial derivatives. After obtaining the single edge pixel in the image, the row and column differentials f1 and f2 are calculated for each pixel. Based on the relationship between f1 and f2, pixel matrices of varying sizes are extracted from the original image to determine the values of the intermediate parameters C, O, S1, S2, and S3. Ultimately, the coefficients of the edge line and the specific position of the sub-pixel are determined.

[0079] As a specific embodiment, detecting the position and direction of the burrs on the wire rope in the background area specifically includes the following steps.

[0080] The background area is calculated using a derivative mask to determine a plurality of single edge pixels in the background area.

[0081] The row difference and column difference of each single edge pixel in the background area are calculated.

[0082] The background area is intercepted based on the row difference and column difference of each single edge pixel in the background area to obtain multiple background area pixel matrices.

[0083] The equation of the background area edge line is determined according to the background area pixel matrix to obtain equations of multiple background area edge lines.

[0084] like Figure 6 As shown, the intersection of the equation for determining the edge straight line of the background area and the straight line where the wire rope is located is the position of the burr, and p1 is the intersection point.

[0085] The normal vector of the equation of the edge line of the background area is the direction of the burr.

[0086] As a specific embodiment, the background area is intercepted based on the row difference and column difference of each single edge pixel in the background area to obtain multiple background area pixel matrices, which specifically includes the following steps.

[0087] It is determined whether the row difference of the single edge pixel in the background area is greater than the column difference of the single edge pixel in the background area to obtain a first determination result.

[0088] If the first judgment result is yes, a background area pixel matrix of size 5×3 is intercepted in the background area.

[0089] If the first judgment result is no, a background area pixel matrix of 3×5 size is intercepted in the background area.

[0090] In the specific implementation, (1) .

[0091] As a specific embodiment, when a 3×5 background area pixel matrix is intercepted in the background area, the equation of the edge line of the 3×5 background area pixel matrix is specifically as follows.

[0092] .

[0093] .

[0094] .

[0095] .

[0096] .

[0097] Where y represents the coordinate of the edge line of the 3×5 background area pixel matrix on the y-axis; Represents the area of all pixels in the first column of the 3×5 background area pixel matrix that are located below the background area edge line; Represents the area of all pixels in the second column of the 3×5 background area pixel matrix that are located below the background area edge line; represents the area of all pixels below the background area edge line in the third column of the 3×5 background area pixel matrix; C represents the mean grayscale value of all pixels below the background area edge line in the 3×5 background area pixel matrix; O represents the mean grayscale value of all pixels above the background area edge line in the 3×5 background area pixel matrix; h represents the side length of the 3×5 background area pixel matrix; a represents the first coefficient; b represents the second coefficient; Represents the grayscale value of the pixel at row i'+1 and column j'-2 in the 3×5 background area pixel matrix; Represents the grayscale value of the pixel at row i' and column j'-2 in the 3×5 background area pixel matrix; Represents the grayscale value of the pixel at row i'+1 and column j'-1 in the 3×5 background area pixel matrix; Represents the grayscale value of the pixel at row i'-1 and column j'+2 in the 3×5 background area pixel matrix; Represents the grayscale value of the pixel at row i' and column j'+2 in the 3×5 background area pixel matrix; represents the grayscale value of the pixel at row i'-1 and column j'+1 in the 3×5 background area pixel matrix; x represents the coordinate of the edge line of the 3×5 background area pixel matrix on the x-axis.

[0098] At this time, the slope of the edge straight line of the 3×5 background area pixel matrix For the sake of convenience, we assume that the slope of the edge line of the 3×5 background area pixel matrix is between (0, 1), and the sub-pixel points to be detected are ,by Pixel as the center, select the background area pixel matrix of 3×5 size, and establish Figure 5 The edge straight line of the 3×5 background area pixel matrix divides the 3×5 background area pixel matrix into two parts, as shown in the following figure: Figure 5 shown.

[0099] According to the method proposed by Trujillo-Pino et al., it is assumed that F1, F2, and F3 are the sum of the grayscale values of each column of the background area pixel matrix of 3×5 size and satisfy the equation: , , .

[0100] Among them, F1 represents the sum of the grayscale values of the pixels in the first column of the background area pixel matrix of 3×5 specifications; F2 represents the sum of the grayscale values of the pixels in the second column of the background area pixel matrix of 3×5 specifications; F3 represents the sum of the grayscale values of the pixels in the third column of the background area pixel matrix of 3×5 specifications.

[0101] . .

[0102] .

[0103] According to the above formula, the expressions of coefficients a and b can be calculated: , .

[0104] Among them, C and O are unknown. Because the edge is affected by optical effects, the pixels at the edge of the structure are not two clear parts, and the grayscale value has a changing process. Therefore, the values of C and O are obtained by the three pixels at the diagonal corners of the edge line through the following method , .

[0105] (2) At this time, the slope of the straight line is , when the line is perpendicular to the horizontal axis, the slope b is infinite. Therefore, for ease of calculation, a horizontal window is selected within this range, and a 5×3 background area pixel matrix is selected. The straight line expression is: x'=a'+b'y'. x' represents the y-axis coordinate of the edge line of the 5×3 background area pixel matrix, a' represents the first coefficient of the equation of the edge line of the 5×3 background area pixel matrix; b' represents the second coefficient of the equation of the edge line of the 5×3 background area pixel matrix; and y' represents the x-axis coordinate of the edge line of the 5×3 background area pixel matrix.

[0106] The method for calculating the parameters of the straight line expression is the same as the above method, so I will not go into details here.

[0107] As a specific embodiment, the burr position and direction of the wire rope in the background area are detected, and then the burr position and direction of the wire rope in the target area are detected, which specifically includes: determining the equations of multiple target area edge lines. The method for determining the equations of the target area edge lines is the same as the method for determining the equations of the background area edge lines. No further details are given here. It is judged whether there are two identical equations in the equations of the multiple target area edge lines to obtain a second judgment result. If the second judgment result is yes, one of the two identical equations is deleted to obtain the equations of the multiple target area edge lines. If the second judgment result is no, it is judged whether the number of equations of the target area edge lines is equal to 2 to obtain a third judgment result. If the third judgment result is yes, it indicates that there are no burrs on the elevator wire rope in the target area. If the third judgment result is no, it indicates that there are burrs on the elevator wire rope in the target area. The equations of the multiple target area edge lines are deleted from the equations of the multiple target area edge lines that are parallel to each other to obtain the equations of the multiple target area edge lines. The burr position and direction are determined based on the equations of the multiple target area edge lines. The method for determining the position and direction of the burr in the target area is the same as that for determining the position and direction of the burr in the background area, and will not be described in detail here.

[0108] Embodiment 2: A non-contact automatic monitoring system for elevator wire rope burrs, the non-contact automatic monitoring system for elevator wire rope burrs comprising: an acquisition module for acquiring an image of the elevator wire rope, a selection module for selecting a wire rope region of interest in the image; a processing module for performing morphological processing on the wire rope region of interest to obtain a morphologically processed wire rope region of interest; a morphological processing model for performing adaptive enhancement processing on the morphologically processed wire rope region of interest to obtain a adaptively enhanced wire rope region of interest; an adaptive enhancement module for performing binarization processing on the adaptively enhanced wire rope region of interest to obtain a binarized wire rope region of interest; a sub-pixel edge detection module for performing sub-pixel edge detection on the binarized wire rope region of interest to determine a target area and a background area; the target area is the wire rope area, and the background area is the burr area; a burr detection module for detecting the position and direction of the burrs on the wire rope in the background area.

[0109] A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed, the non-contact automatic monitoring method for elevator wire rope burrs as described in Example 1 is implemented.

[0110] The present invention is simple to operate, easy to use, and highly efficient, providing more accurate results than traditional manual inspection methods. By performing a series of processing on contactless images, the present invention can determine the location and direction of burrs on the wire rope. This solves the problems of on-site inspection difficulties and missed detections in existing elevator wire rope burr inspection methods, reduces the impact of the operator's technical level on the inspection results, enables automated inspection, shortens inspection time, and improves work efficiency.

[0111] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0112] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only intended to help understand the method and core concept of the present invention. At the same time, those skilled in the art will find that the specific implementation methods and application scopes may vary based on the concept of the present invention. In summary, the contents of this specification should not be construed as limiting the present invention.

Claims

1. A non-contact automatic monitoring method for elevator wire rope burrs, characterized in that: The non-contact automatic monitoring method for elevator wire rope burrs includes: Acquire images of elevator ropes; Selecting a wire rope region of interest in the image; Performing morphological processing on the steel wire rope region of interest to obtain a morphologically processed steel wire rope region of interest; Performing adaptive enhancement processing on the wire rope region of interest after morphological processing to obtain the wire rope region of interest after adaptive enhancement processing; Binarization is performed on the wire rope region of interest after the adaptive enhancement process to obtain a binarized wire rope region of interest; Sub-pixel edge detection is performed on the wire rope region of interest after the binarization process to determine the target region and the background region; the target region is the wire rope region, and the background region is the burr region; Detect the position and direction of burrs on the wire rope in the background area; The adaptive enhancement processing is performed on the wire rope region of interest after morphological processing to obtain the wire rope region of interest after adaptive enhancement processing, specifically including: Constructing a global two-dimensional histogram of the morphologically processed wire rope region of interest; an element f(i, j) in the global two-dimensional histogram represents the number of occurrences of a two-dimensional vector (i, j), where the two-dimensional vector (i, j) is composed of the grayscale value i of the pixel and the neighborhood mean j of the pixel; Based on the global two-dimensional histogram, the Otsu method is used to obtain a first optimal threshold and a second optimal threshold; the first optimal threshold is the optimal threshold of the wire rope region of interest after morphological processing in the grayscale dimension, and the second optimal threshold is the optimal threshold of the wire rope region of interest after morphological processing in the neighborhood grayscale mean dimension; The global two-dimensional histogram is divided into a first region, a second region, a third region, and a fourth region based on the first optimal threshold and the second optimal threshold; the first region is a region whose abscissa is in the interval (0, s') and whose ordinate is in the interval (0, t'), the second region is a region whose abscissa is in the interval (0, s') and whose ordinate is in the interval (t', L-1), the third region is a region whose abscissa is in the interval (s', L-1) and whose ordinate is in the interval (t', L-1), and the fourth region is a region whose abscissa is in the interval (s', L-1) and whose ordinate is in the interval (0, t'); s' represents the first optimal threshold; t' represents the second optimal threshold; and L represents the number of gray levels; The maximum inter-class variance method is used to calculate the global thresholds of the second and fourth regions; Dividing the second area into a noise area and a non-noise area based on the global threshold of the second area; the noise area is an area with a horizontal coordinate in the interval (0, s') and a vertical coordinate in the interval (t', t1), and the non-noise area is an area with a horizontal coordinate in the interval (0, s') and a vertical coordinate in the interval (t1, L-1); t1 represents the global threshold of the second area; The fourth region is divided into an edge region and a non-edge region based on a global threshold of the fourth region; the edge region is a region whose abscissa is in the interval (s', s1) and whose ordinate is in the interval (0, t'); the non-noise region is a region whose abscissa is in the interval (s1, L-1) and whose ordinate is in the interval (0, t'); s1 represents the global threshold of the fourth region; Performing median processing on the grayscale values of pixels in the noise area of the second area to obtain a second area after the median processing; Based on the grayscale value range of the first region and the grayscale value range of the edge region, mapping the grayscale values of pixels in the edge region of the fourth region to obtain a mapped fourth region; The first region, the second region after the median processing, the third region and the fourth region after the mapping processing are used as the wire rope interest region after the adaptive enhancement processing.

2. The non-contact automatic monitoring method for elevator wire rope burrs according to claim 1 is characterized in that: Based on the global two-dimensional histogram, the first optimal threshold and the second optimal threshold are obtained using the Otsu method, specifically including: Based on the global two-dimensional histogram, a discrete measurement function is obtained using the Otsu method; The first threshold and the second threshold when the discrete measure function reaches its maximum value are taken as the first optimal threshold and the second optimal threshold.

3. The non-contact automatic monitoring method for elevator wire rope burrs according to claim 2, characterized in that: The discrete measure function is: S B (s,t)=w0[(μ0-μ T ) 2 +(μ'0-μ' T ) 2 ]+w1[(μ1-μ T ) 2 +(μ′1-μ′ T ) 2 ]; Among them, S B (s, t) represents a discrete measure function; μ0 represents the mean of the grayscale values of all pixels in the first region divided by the first threshold s and the second threshold t; μ′0 represents the mean of the neighborhood means of all pixels in the first region divided by the first threshold s and the second threshold t; μ1 represents the mean of the grayscale values of all pixels in the third region divided by the first threshold s and the second threshold t; μ′1 represents the mean of the neighborhood means of all pixels in the third region divided by the first threshold s and the second threshold t; μ T Represents the mean grayscale value of all pixels in the global two-dimensional histogram; μ′ T represents the mean of the neighborhood means of all pixels in the global two-dimensional histogram; w0 represents the percentage of the first area divided by the first threshold s and the second threshold t in the global two-dimensional histogram; w1 represents the percentage of the third area divided by the first threshold s and the second threshold t in the global two-dimensional histogram.

4. The non-contact automatic monitoring method for elevator wire rope burrs according to claim 1, characterized in that: Detect the location and direction of burrs on the wire rope in the background area, including: Calculating the background area using a derivative mask to determine a plurality of single edge pixels in the background area; Calculating row differences and column differences of each single edge pixel in the background area; intercepting the background area based on the row difference and column difference of each single edge pixel in the background area to obtain a plurality of background area pixel matrices; Determining an equation of a background area edge line according to the background area pixel matrix to obtain equations of multiple background area edge lines; Determine the intersection of the equation of the edge line of the background area and the line where the wire rope is located as the location of the burr; The normal vector of the equation of the edge line of the background area is the direction of the burr.

5. The non-contact automatic monitoring method for elevator wire rope burrs according to claim 4, characterized in that: The background area is intercepted based on the row difference and column difference of each single edge pixel in the background area to obtain multiple background area pixel matrices, specifically including: Determine whether the row difference of the single edge pixel in the background area is greater than the column difference of the single edge pixel in the background area, and obtain a first determination result; If the first judgment result is yes, intercepting a 5×3 background area pixel matrix in the background area; If the first judgment result is no, a background area pixel matrix of 3×5 size is intercepted in the background area.

6. The non-contact automatic monitoring method for elevator wire rope burrs according to claim 5, characterized in that: When a 3×5 background area pixel matrix is intercepted in the background area, the equation of the edge straight line of the 3×5 background area pixel matrix is: y=a+bx; Wherein, y represents the coordinate of the edge line of the 3×5 background area pixel matrix on the y-axis; S1 represents the area of all pixels below the background area edge line in the first column of the 3×5 background area pixel matrix; S2 represents the area of all pixels below the background area edge line in the second column of the 3×5 background area pixel matrix; S3 represents the area of all pixels below the background area edge line in the third column of the 3×5 background area pixel matrix; C represents the mean grayscale value of all pixels below the background area edge line in the 3×5 background area pixel matrix; O represents the mean grayscale value of all pixels above the background area edge line in the 3×5 background area pixel matrix; h represents the side length of the 3×5 background area pixel matrix; a represents the first coefficient; b represents the second coefficient; F i′+1，j′-2 Indicates the grayscale value of the pixel at row i'+1 and column j'-2 in the 3×5 background area pixel matrix; F i′，j′-2 Indicates the grayscale value of the pixel at row i' and column j'-2 in the 3×5 background area pixel matrix; F i′+1，j′-1 Indicates the grayscale value of the pixel at row i'+1 and column j'-1 in the 3×5 background area pixel matrix; F i′-1，j′+2 Indicates the grayscale value of the pixel at row i'-1 and column j'+2 in the 3×5 background area pixel matrix; F i′，j′+2 Indicates the grayscale value of the pixel at row i' and column j'+2 in the 3×5 background area pixel matrix; F i′-1,j′+1 represents the grayscale value of the pixel at row i'-1 and column j'+1 in the 3×5 background area pixel matrix; x represents the coordinate of the edge line of the 3×5 background area pixel matrix on the x-axis.

7. The non-contact automatic monitoring method for elevator wire rope burrs according to claim 1, characterized in that: Detect the position and direction of burrs on the wire rope in the background area, followed by: Detect the location and direction of burrs on the wire rope within the target area.

8. A non-contact automatic monitoring system for elevator wire rope burrs, characterized in that: The non-contact automatic monitoring system for elevator wire rope burrs includes: An acquisition module, used for acquiring an image of the elevator wire rope; A selection module, configured to select a wire rope region of interest in the image; a processing module, configured to perform morphological processing on the steel wire rope region of interest to obtain the morphologically processed steel wire rope region of interest; A morphological processing model is used to perform adaptive enhancement processing on the wire rope region of interest after morphological processing to obtain the wire rope region of interest after adaptive enhancement processing; An adaptive enhancement module is used to perform a binarization process on the wire rope region of interest after the adaptive enhancement process to obtain a binarized wire rope region of interest; A sub-pixel edge detection module is used to perform sub-pixel edge detection on the wire rope region of interest after the binarization process to determine the target region and the background region; the target region is the wire rope region, and the background region is the burr region; The burr detection module is used to detect the position and direction of burrs on the wire rope in the background area; The adaptive enhancement processing is performed on the wire rope region of interest after morphological processing to obtain the wire rope region of interest after adaptive enhancement processing, specifically including: Constructing a global two-dimensional histogram of the morphologically processed wire rope region of interest; an element f(i, j) in the global two-dimensional histogram represents the number of occurrences of a two-dimensional vector (i, j), where the two-dimensional vector (i, j) is composed of the grayscale value i of the pixel and the neighborhood mean j of the pixel; Based on the global two-dimensional histogram, the Otsu method is used to obtain a first optimal threshold and a second optimal threshold; the first optimal threshold is the optimal threshold of the wire rope region of interest after morphological processing in the grayscale dimension, and the second optimal threshold is the optimal threshold of the wire rope region of interest after morphological processing in the neighborhood grayscale mean dimension; The global two-dimensional histogram is divided into a first region, a second region, a third region, and a fourth region based on the first optimal threshold and the second optimal threshold; the first region is a region whose abscissa is in the interval (0, s') and whose ordinate is in the interval (0, t'), the second region is a region whose abscissa is in the interval (0, s') and whose ordinate is in the interval (t', L-1), the third region is a region whose abscissa is in the interval (s', L-1) and whose ordinate is in the interval (t', L-1), and the fourth region is a region whose abscissa is in the interval (s', L-1) and whose ordinate is in the interval (0, t'); s' represents the first optimal threshold; t' represents the second optimal threshold; and L represents the number of gray levels; The maximum inter-class variance method is used to calculate the global thresholds of the second and fourth regions; Dividing the second area into a noise area and a non-noise area based on the global threshold of the second area; the noise area is an area with a horizontal coordinate in the interval (0, s') and a vertical coordinate in the interval (t', t1), and the non-noise area is an area with a horizontal coordinate in the interval (0, s') and a vertical coordinate in the interval (t1, L-1); t1 represents the global threshold of the second area; The fourth region is divided into an edge region and a non-edge region based on a global threshold of the fourth region; the edge region is a region whose abscissa is in the interval (s', s1) and whose ordinate is in the interval (0, t'); the non-noise region is a region whose abscissa is in the interval (s1, L-1) and whose ordinate is in the interval (0, t'); s1 represents the global threshold of the fourth region; Performing median processing on the grayscale values of pixels in the noise area of the second area to obtain a second area after the median processing; Based on the grayscale value range of the first region and the grayscale value range of the edge region, mapping the grayscale values of pixels in the edge region of the fourth region to obtain a mapped fourth region; The first region, the second region after the median processing, the third region and the fourth region after the mapping processing are used as the wire rope interest region after the adaptive enhancement processing.

9. A computer-readable storage medium, characterized in that The storage medium stores a computer program, which implements the method according to any one of claims 1 to 7 when executed.

Images