Belt conveyor belt detection method

Abstract

The invention relates to the technical field of image processing, in particular to a belt conveyor belt detection method which comprises the steps that an image of a conveyor belt is collected, and a pit image of the image is obtained through a pit recognition network; according to the fact that any pit in the current-frame pit image is a target pit, the matching index of the target pit and any pit in the next-frame pit image is calculated, the target pit in the continuous-frame pit image is determined, the matching index of the two pits is in positive correlation with the similarity between the surrounding pits of the two pits, and the matching index of the two pits is not in positive correlation with the similarity between the surrounding pits of the two pits. The similarity is in positive correlation with the similarity between the two pits; and calculating the clear degree of each target pit in the continuous frames of pit images. According to the method, the abrasion coefficient is calculated, so that the abrasion of the conveying belt can be quantitatively analyzed, and safety judgment is performed according to a set threshold value. Therefore, the wear detection does not depend on subjective evaluation, but judges whether an alarm is needed or not through a clear standard, so that the reliability of decision making is enhanced.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and more specifically, to a belt conveyor belt detection method. Background Art

[0002] Belt conveyors are widely used in many industries such as mining, manufacturing and logistics for efficient material transportation. As its core component, the condition of the conveyor belt is directly related to the reliability and safety of the entire system. In coal mining, conveyor belts are an indispensable and important equipment in coal mines, thermal power plants and coal transportation systems. It is mainly used in the mining, transportation and loading and unloading of coal, and is one of the key links in modern coal production. With the continuous development of the coal industry, the role of conveyor belts as a means of transportation has become more and more significant. Its main function is to transport coal from the mining site to the processing facility or to other means of transportation, saving a lot of labor and time costs. However, long-term operation and harsh working environment make coal conveyor belts prone to wear, aging and damage, which affects production efficiency and safety.

[0003] In the related technology, for example, the Chinese patent application document with publication number CN112598720A discloses a conveyor belt wear status detection method and system based on binocular photography. According to the surface contour feature information, the uniformity of the contour line distribution spacing of the surface contour is determined, and then the distribution state of the contour lines of the surface contour is determined according to the uniformity of the contour line distribution spacing at each pixel point; finally, the surface wear position area of ​​the conveyor belt is determined according to the uniformity of the contour line distribution spacing of the surface contour, so as to accurately locate the surface wear position area of ​​the conveyor belt.

[0004] As coal falls onto the conveyor belt, the impact of the coal on the conveyor belt will cause pits of varying degrees to form on the conveyor belt. However, since the size of the pits is uncertain and some smaller pits may be inaccurately identified, it is impossible to accurately detect the wear of the conveyor belt. Summary of the invention

[0005] The present invention provides a belt conveyor belt detection method, aiming to solve the problem in the related technology that the impact of coal on the conveyor belt will cause the conveyor belt to be dented to varying degrees, but since the size of the pits is uncertain, some smaller pits may be inaccurately identified, and thus the wear of the conveyor belt cannot be accurately detected.

[0006] The present invention provides a belt conveyor belt detection method, comprising: collecting an image of a conveyor belt, and obtaining a pit image of the image by using a pit recognition network; according to any pit in the pit image of a current frame being a target pit, calculating a matching index between the target pit and any pit in the pit image of a next frame, so as to determine the target pit in the pit image of a continuous frame, wherein the matching index of the two pits is positively correlated with the similarity between the surrounding pits of the two pits, and is also positively correlated with the similarity between the two pits; calculating the clarity of each target pit in the pit image of a continuous frame, and selecting a pit image corresponding to the highest clarity of the pit as the target image, wherein the clarity of the pit reflects the degree of confusion of the pixel value in the pit and the difference in the pixel value in the pit; calculating the wear coefficient of the target pit in the target image, and performing wear detection on the conveyor belt according to the size of the wear coefficient, wherein the wear coefficient of the target pit is positively correlated with the area of ​​the target pit, and is also positively correlated with the ratio of the minimum pixel value of the target pit to the average pixel value of a normal area. By combining image processing technology and deep learning networks, the pit recognition network is used to accurately extract the pit features on the conveyor belt. By analyzing the pit image and calculating the matching index, clarity and other parameters of the pit, the accuracy of the detection of the wear condition of the conveyor belt can be effectively improved. In summary, this method can efficiently, accurately and real-timely detect the wear condition of the coal conveyor belt, improve safety, and reduce manual intervention.

[0007] Furthermore, the conveyor belt is tested for wear according to the size of the wear coefficient, including: in response to the wear coefficient being greater than the wear evaluation threshold, an alarm is issued; in response to the wear coefficient being less than or equal to the wear evaluation threshold, no alarm is issued. By setting the wear evaluation threshold, an alarm can be triggered when the wear coefficient is greater than the wear evaluation threshold. In this way, potential safety hazards can be discovered in a timely manner, reducing the risk of accidents.

[0008] Furthermore, the calculation formula of the wear coefficient of the target pothole in the target image is: ; In the formula, is the wear coefficient of the target pothole in the target image, is the number of pixels of the target holes in the target image, is the pixel value sequence of the target hole in the target image, is the mean pixel value of the non-hole area in the target image, is a hyperbolic tangent function. The calculation of the wear coefficient takes into account multiple factors such as the area of ​​the target pothole and the difference in pixel values, which can fully reflect the wear condition of the conveyor belt.

[0009] Furthermore, calculating the wear coefficient of the target pothole also includes: obtaining the difference between the pixel value of the target pothole and the pixel value of the non-pothole area in the target image to obtain a pothole difference map, performing k-means clustering on the pothole difference map, determining the number of clusters using the silhouette coefficient method, and calculating the wear coefficient of the target pothole based on the number of clusters. The calculation formula is: ; In the formula, is the wear coefficient of the target pothole, is the pixel value sequence composed of the center pixels of all clusters. is the mean pixel value of the non-hole area in the target image, is the number of cluster centers of all clusters, is the hyperbolic tangent function. Furthermore, the clarity of each target pothole in the continuous frame pothole image is calculated, and the calculation formula is: ; In the formula, For the The clarity of the target pit in the frame pit image, For the The pixel value of the target hole in the frame hole image is The proportion of For the The pixel value sequence of the target pothole in the frame pothole image.

[0010] Furthermore, the matching index between the target pothole and any pothole in the next pothole image is calculated, and the calculation formula is: ; In the formula, The target pothole and the next frame of the pothole image The matching index of the holes, is the pixel value matrix of the target pothole, For the next frame The pixel value matrix of the pits, For the next frame The pixel value matrix of the pits, The target pothole is surrounded by The position vector is composed of the geometric center of each pit pointing to the geometric center of the target pit. In the next frame of the pothole image, Around the pothole The geometric center of the pit points to The position vector composed of the geometric centers of the pits, and Both are similarity calculations.

[0011] Furthermore, the training process of the pothole recognition network includes: marking the pothole pixels in the conveyor belt image as 1 and the non-pothole pixels as 0 to obtain a 0-1 marked image; inputting the 0-1 marked image into the semantic segmentation network training; completing the training in response to the loss function being less than a preset value or reaching a preset number of training times, to obtain a pothole recognition network. By training the pothole recognition network, the system can automatically extract effective features from the image. And with the continuous optimization of network training, the recognition accuracy and adaptability of the system will gradually improve, enhancing its feasibility for long-term use.

[0012] Furthermore, a pothole image of the image is obtained by using a pothole recognition network, including: inputting the image of the conveyor belt into the pothole recognition network to obtain a 0-1 labeled image; and obtaining a pothole image based on multiplying the 0-1 labeled image matrix with the image matrix of the conveyor belt.

[0013] Furthermore, the image of the conveyor belt is collected by a depth camera, wherein the pixels in the image contain grayscale information and depth information. By combining depth information and grayscale value, not only the shape of the potholes can be detected, but also the depth and size of the potholes can be more accurately identified, thus improving the adaptability in complex working environments.

[0014] Further, obtaining each pothole in the pothole image includes: obtaining each connected domain in the pothole image using a connected domain analysis method, wherein one connected domain is a pothole.

[0015] Beneficial Effects (1) The deep learning algorithm can automatically identify pothole images on the conveyor belt, efficiently extract pothole features from complex images, reduce manual intervention, and improve the degree of automation of detection.

[0016] (ii) By calculating the clarity of each frame of the target pothole image, the quality of the pothole image can be judged to ensure that the clearest image is selected as the final target image. The clarity is not only related to the difference and degree of confusion of the pixels in the pothole, but can also effectively distinguish relatively blurred or low-quality images to avoid misinterpretation of the image. By selecting the image with the highest clarity as the target image, it can ensure that the subsequent wear coefficient calculation is based on high-quality data, thereby improving the accuracy of the detection results.

[0017] (III) By calculating the wear coefficient, the wear of the conveyor belt can be quantitatively analyzed and safety judgment can be made based on the set threshold. This makes wear detection not rely on subjective evaluation, but judges whether an alarm is needed through clear standards, thereby enhancing the reliability of decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1FIG. 4 is a flow chart schematically illustrating a method for calculating a wear coefficient of a conveyor belt according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0020] like Figure 1 As shown, S101: acquiring a pothole image of the conveyor belt.

[0021] Specifically, the surface of the conveyor belt is photographed by an image acquisition device to obtain an image of the conveyor belt, and the potholes in the image are identified using a semantic segmentation network (CNN) to extract the pothole image. The image acquisition device is a depth camera, so that each pixel in the image contains depth information and gray value information, which is convenient for subsequent wear detection of the conveyor belt.

[0022] In one embodiment, a semantic segmentation network (CNN) is used to identify potholes in an image and extract a pothole image, including: training a pothole recognition network, inputting the image of a conveyor belt into the trained pothole recognition network, and obtaining a pothole image. The process of training the pothole recognition network is as follows: marking the pothole pixels in the image of the conveyor belt as 1 and the non-pothole pixels as 0 to obtain a 0-1 marked image; inputting the 0-1 marked image into the semantic segmentation network for training; completing the training in response to the loss function being less than a preset value or reaching a preset number of training times, and obtaining a pothole recognition network. Then, the image of the conveyor belt is input into the pothole recognition network to obtain a 0-1 marked image, and the 0-1 marked image matrix is ​​multiplied by the image matrix of the conveyor belt to obtain a pothole image. At this point, the pothole image corresponding to each frame of the conveyor belt image can be obtained.

[0023] By training pothole images through a deep learning network, the algorithm can be continuously optimized to improve the system's adaptability in different environments. In different working environments and on different types of conveyor belts, the system can automatically adjust the recognition and detection strategies based on training data and real-time images, maintain high detection accuracy, reduce manual operations, and reduce human omissions and errors, thereby improving the reliability of the overall system.

[0024] In one embodiment, after obtaining the pothole image corresponding to each frame of the conveyor belt image, a connected domain analysis method is used to obtain each connected domain in the pothole image, where one connected domain represents a pothole. Thus, each pothole in the pothole image can be obtained.

[0025] It should be noted that because the conveyor belt runs at a uniform speed in actual operation, that is, fixed-interval conveyor belt surface images are collected at equal time intervals, and continuous frame pothole images can be obtained. However, due to the uniform motion of the conveyor belt, some of the collected images are not clear. Therefore, it is necessary to screen the pothole images of continuous frames to select the pothole images with the highest clarity, to ensure that the subsequent wear coefficient calculation is based on high-quality data, thereby improving the accuracy of the analysis of the degree of conveyor belt wear.

[0026] S102: Obtain the position of any pothole in the pothole images of consecutive frames.

[0027] Specifically, according to any pothole in the current frame pothole image being the target pothole, the matching index between the target pothole and any pothole in the next frame pothole image is calculated, and the target pothole in the continuous frame pothole image is determined according to the size of the matching index, wherein the matching index of the two potholes is positively correlated with the similarity between the surrounding potholes of the two potholes, and is also positively correlated with the similarity between the two potholes. Specifically, the matching index of all potholes at the edge of the next frame pothole image is obtained, and when the matching index reaches the maximum value, it indicates that the pothole and the target pothole are the same pothole. At this point, the position of the target pothole in the current frame pothole image on the next frame pothole image can be obtained, and then the position of the target pothole on the continuous frame pothole image can be determined.

[0028] In one embodiment, the matching index between the target pothole and any pothole in the next pothole image is calculated using the following formula: ; In the formula, The target pothole and the next frame of the pothole image The matching index of the holes, is the pixel value matrix of the target pothole, For the next frame The pixel value matrix of the pits, For the next frame The pixel value matrix of the pits, The target pothole is surrounded by The position vector is composed of the geometric center of each pit pointing to the geometric center of the target pit. In the next frame of the pothole image, Around the pothole The geometric center of the pit points to The position vector composed of the geometric centers of the pits, and Both are similarity calculations.

[0029] in, It is the similarity between the pixel value matrix of the kth pit in the next pit image and the target pit. The closer this value is to 1, the more similar the two pits are. It is the similarity between the four nearest pits around the target pit and the kth pit in the next pit image and the four nearest pits around the target pit in the current pit image, including the similarity comparison of the position and the pixel value matrix in the pit. The larger the value, the more likely it is that there is a highly similar pit distribution around the target pit and the kth pit in the next pit image, and the target pit and the kth pit in the next pit image are the same pit.

[0030] By matching the pits using the above method, we can accurately find the position of the target pit in the pit images of adjacent frames, and effectively avoid the mismatch of the target pit caused by similar pits through the position and internal pixel value distribution of adjacent pits, thereby ensuring the accuracy of the subsequent calculation of the clarity of the pits.

[0031] It should be noted that the user can appropriately increase or decrease the number of adjacent pits calculated according to the actual application scenario. In principle, the finer the coal, the more adjacent pits should be involved in the calculation. The reason is that the coal unloading device is located above the conveyor belt. The coal can fall onto the conveyor belt through the unloading device and cause impact on the conveyor belt. Because the smaller the coal, the more times the conveyor belt will be impacted, and the more similar the impact force, the size of the pits will be roughly the same. Therefore, more pit distribution features are needed to improve the accuracy of matching.

[0032] S103: Calculate the clarity of each target pothole in the pothole images of consecutive frames, and select the target image.

[0033] Specifically, the clarity of each target pit in the continuous frame pit images is calculated, and the continuous frame pit images all contain complete target pits, and the pit image corresponding to the highest pit clarity is selected as the target image, wherein the clarity of the pit reflects the degree of confusion of the pixel values ​​in the pit and the difference in the pixel values ​​in the pit.

[0034] In one embodiment, the clarity of each target pothole in the pothole images of consecutive frames is calculated using the following calculation formula: ; In the formula, For the The clarity of the target pit in the frame pit image, For the The pixel value of the target hole in the frame hole image is The proportion of For the The pixel value sequence of the target pothole in the frame pothole image.

[0035] in, is the degree of confusion of the target pit pixel value in any frame pit image. The larger the value, the more pixel value information the target pit image contains in any frame pit image. The larger the value is, the greater the difference in pixel value levels contained in the target pothole image in the pothole image of any frame is, and the clearer the pothole is. Thus, the clarity of the target pothole in each consecutive pothole image can be obtained, and the pothole image corresponding to the highest pothole clarity is selected as the target image. By obtaining the clearest pothole image corresponding to the target pothole in a continuous shooting process, the accuracy of the subsequent calculation of the wear coefficient of the pothole is improved.

[0036] S104: Calculate the wear evaluation of the target pothole.

[0037] Specifically, the wear coefficient of the target pothole in the target image is calculated, and the conveyor belt is tested for wear according to the size of the wear coefficient, wherein the wear coefficient of the target pothole is positively correlated with the area of ​​the target pothole, and is also positively correlated with the ratio of the minimum pixel value of the target pothole to the average pixel value of the normal area.

[0038] In one embodiment, the calculation formula of the wear coefficient of the target pothole in the target image is: ; In the formula, is the wear coefficient of the target pothole in the target image, is the number of pixels of the target holes in the target image, is the pixel value sequence of the target hole in the target image, Represents the minimum value of the pixel in the pixel value sequence of the target pit in the target image. is the mean pixel value of the non-hole area in the target image, is the hyperbolic tangent function. Among them, It is the comparison between the minimum value of the pixel value sequence of the clearest target pothole and the mean value of the pixel value in the non-pothole area. The larger the value, the more serious the wear of the clearest target pothole. It is an expression of the size of the pothole. The larger the value, the larger the pothole area. It should be noted that when calculating the wear coefficient, multiple factors such as the area of ​​the target pothole and the difference in pixel values ​​are taken into account, which can fully reflect the wear of the conveyor belt.

[0039] In another embodiment, another method for calculating the wear coefficient of the target pothole in the target image is provided, and the specific steps are as follows: subtract the mean of the pixel value of the target pothole in the target image from the pixel value of the non-pothole area to obtain a pothole difference map, and perform k-means clustering on the pothole difference map, wherein the initial clustering cluster k is set to 1, and the contour coefficient value is obtained by using the contour coefficient method, and a contour coefficient value is obtained each time a cluster is added, and the contour coefficient threshold is set to 0.5. When the contour coefficient value is greater than the contour coefficient threshold, the optimal number of clustering clusters is obtained. It should be noted that the more clustering clusters there are, the worse the quality of the pothole. Because a clustering cluster means that the cluster center has received greater wear. The wear coefficient of the target pothole is calculated according to the above method, and the calculation formula is: .in, is the wear coefficient of the target pothole, is the pixel value sequence composed of the center pixels of all clusters. is the mean pixel value of the non-hole area in the target image, is the number of cluster centers of all clusters, is the hyperbolic tangent function. The wear degree of the worst position point in the target pothole can be found as the overall wear degree of the target pothole, while the number of wear points in the target pothole is taken into account as the overall quality of the pothole.

[0040] S105: Perform wear detection on the conveyor belt according to the size of the wear coefficient.

[0041] In one embodiment, if the wear coefficient is greater than the wear evaluation threshold, it means that the wear at the target pothole position no longer meets the use requirements, and an alarm prompt is issued to remind relevant staff that repairs or replacements are required; if the wear coefficient is less than or equal to the wear evaluation threshold, no alarm is issued and the conveyor belt can continue to be used. In this embodiment, the empirical value of the wear evaluation threshold is 0.8. In other embodiments, the empirical value of the wear evaluation threshold can be 0.85 or 0.78, etc., which can be adjusted according to the specific implementation situation. By setting the wear evaluation threshold, when the wear coefficient is greater than the wear evaluation threshold, an alarm prompt can be triggered, which can timely discover potential safety hazards and reduce the risk of accidents.

[0042] Through the above steps, image processing technology and deep learning network are combined to accurately extract the characteristics of potholes on the conveyor belt using the pothole recognition network. By analyzing the pothole image and calculating the matching index, clarity and other parameters of the pothole, the accuracy of the detection of the wear condition of the conveyor belt can be effectively improved. In summary, this method can efficiently, accurately and real-time detect the wear condition of the coal conveyor belt, improve safety, and reduce manual intervention.

[0043] The above-mentioned embodiments only express several implementation methods of the present invention, and the description is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent application. It should be pointed out that for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A belt conveyor belt detection method, characterized in that: include: Collecting an image of the conveyor belt, and obtaining a pothole image of the image using a pothole recognition network; According to any pothole in the pothole image of the current frame being the target pothole, a matching index between the target pothole and any pothole in the pothole image of the next frame is calculated to determine the target pothole in the pothole images of the consecutive frames, wherein the matching index of the two potholes is positively correlated with the similarity of the pothole distribution between the surrounding potholes of the two potholes and is also positively correlated with the similarity between the two potholes; Calculating the clarity of each target pit in the continuous frame pit images, wherein the continuous frame pit images all contain complete target pits, and selecting the pit image corresponding to the highest pit clarity as the target image, wherein the pit clarity reflects the degree of confusion of pixel values ​​in the pit and the size of the difference of pixel values ​​in the pit; The wear coefficient of the target pothole in the target image is calculated, and the conveyor belt is tested for wear according to the size of the wear coefficient, wherein the wear coefficient of the target pothole is positively correlated with the area of ​​the target pothole, and is also positively correlated with the ratio of the minimum pixel value of the target pothole to the average pixel value of the normal area.

2. The belt conveyor belt detection method according to claim 1, characterized in that: According to the size of the wear coefficient, the conveyor belt is subjected to wear detection, including: In response to the wear coefficient being greater than a wear evaluation threshold, an alarm is issued; In response to the wear coefficient being less than or equal to the wear evaluation threshold, no alarm is issued.

3. The belt conveyor belt detection method according to claim 1, characterized in that: The calculation formula of the wear coefficient of the target pothole in the target image is: ； In the formula, is the wear coefficient of the target pothole in the target image, is the number of pixels of the target holes in the target image, is the pixel value sequence of the target hole in the target image, is the mean pixel value of the non-hole area in the target image, is the hyperbolic tangent function.

4. The belt conveyor belt detection method according to claim 1, characterized in that: Calculate the wear coefficient of the target pothole, including: The pixel value of the target pothole in the target image is obtained by subtracting the pixel value of the non-pothole area to obtain the pothole difference map, and the pothole difference map is clustered by k-means. The number of clusters is determined by the silhouette coefficient method, and the wear coefficient of the target pothole is calculated based on the number of clusters. The calculation formula is: ； In the formula, is the wear coefficient of the target pothole, is the pixel value sequence composed of the center pixels of all clusters. is the mean pixel value of the non-hole area in the target image, is the number of cluster centers of all clusters, is the hyperbolic tangent function.

5. The belt conveyor belt detection method according to claim 1, characterized in that: Calculate the clarity of each target pit in the continuous frame pit image, the calculation formula is: ； In the formula, For the The clarity of the target pit in the frame pit image, For the The pixel value of the target hole in the frame hole image is The proportion of For the The pixel value sequence of the target pothole in the frame pothole image.

6. The belt conveyor belt detection method according to claim 1, characterized in that: Calculate the matching index between the target pothole and any pothole in the next pothole image. The calculation formula is: ； In the formula, The target pothole and the next frame of the pothole image The matching index of the holes, is the pixel value matrix of the target pothole, For the next frame The pixel value matrix of the pits, For the next frame The pixel value matrix of the pits, The target pothole is surrounded by The position vector is composed of the geometric center of each pit pointing to the geometric center of the target pit. In the next frame of the pothole image, Around the pothole The geometric center of the pit points to The position vector composed of the geometric centers of the pits, and Both are similarity calculations.

7. The belt conveyor belt detection method according to claim 1, characterized in that: The training process of the pothole recognition network includes: The pothole pixels in the conveyor belt image are marked as 1, and the non-pothole pixels are marked as 0, to obtain a 0-1 marked image; Input the 0-1 labeled image into a semantic segmentation network for training; In response to the loss function being less than a preset value or reaching a preset number of training times, the training is completed to obtain a pothole recognition network.

8. The belt conveyor belt detection method according to claim 1, characterized in that: Obtaining a pothole image of the image using a pothole recognition network includes: Inputting the image of the conveyor belt into a pothole recognition network to obtain a 0-1 labeled image; The pothole image is obtained by multiplying the 0-1 labeled image matrix with the image matrix of the conveyor belt.

9. The belt conveyor belt detection method according to claim 1, characterized in that: The image of the conveyor belt is captured by a depth camera, wherein the pixels in the image contain grayscale value information and depth information.

10. The belt conveyor belt detection method according to claim 1, characterized in that: Get each pothole in the pothole image, including: The connected domain analysis method is used to obtain the connected domains in the pothole image, where one connected domain is a pothole.

Images