Method and device for monitoring deviation degree of conveying belt in real time, storage medium and processor

Abstract

The embodiment of the invention provides a real-time conveyor belt deviation degree monitoring method and device, a storage medium and a processor, and belongs to the technical field of image processing. The method comprises the following steps: acquiring a historical image set of normal operation and off-tracking operation of the conveyor belt; marking the carrier rollers on the left side and the right side closest to the camera in each historical image in the historical image set to obtain an image set with a label; carrying out enhancement processing on the image set with the label and then training YOLOv11 to obtain a target segmentation model; acquiring a real-time image of the operation of the conveying belt, and inputting the real-time image into the target segmentation model to obtain segmentation feature images of the carrier rollers on the left and right sides closest to the camera; and according to the segmentation feature images of the carrier rollers on the left side and the right side closest to the camera and a preset formula, the real-time deviation amount of the conveying belt is obtained through calculation. The method does not need to preset edge points of the conveying belt or recognize the edge of the conveying belt, is good in deviation recognition effect and high in efficiency, reduces manual intervention, and improves the working efficiency and safety.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a method, device, storage medium and processor for real-time monitoring of the deviation degree of a conveyor belt. Background Art

[0002] Belt conveyors are widely used in the coal conveying systems of coal-fired power plants. During actual operation, due to installation problems, belt damage, deviation of the material unloading point, and heavy material load, the conveyor belt often deviates, seriously affecting production efficiency and posing great potential safety hazards to industrial production. First, the deviation phenomenon can cause the system to stop due to faults, affecting production operation efficiency; second, in the case of deviation, the conveyor belt will increase the axial force borne by the rollers and idlers, causing the rollers to shift axially and the idler bearings to be damaged. At the same time, when the deviation is severe, it will lead to safety hazards such as transverse tearing of the conveyor belt; finally, when the conveyor belt deviates, it will cause the transported materials to spill and generate dust, polluting the environment. Therefore, the monitoring of conveyor belt deviation is particularly important. Existing technologies obtain the deviation situation of the conveyor belt by judging the position of the conveyor belt or the positional relationship between the conveyor belt and the idlers. These methods all require extracting the edge of the conveyor belt. However, in practical applications, it is easily affected by environmental brightness, belt reflection, light spots generated by lighting equipment, and harsh working environments, and the effect of identifying and extracting the edge of the conveyor belt is often unsatisfactory. And these methods rely on preset conveyor belt edge points and need to calibrate these preset points regularly, which makes the effect of identifying and extracting the edge of the conveyor belt sensitive to the deviation and disturbance of the camera, and the algorithm deployment is also relatively cumbersome. Summary of the Invention

[0003] The object of the present invention is to provide a method, device, storage medium and processor for real-time monitoring of the deviation degree of a conveyor belt, which are used to realize the real-time monitoring of the conveyor belt and correct the deviation of the conveyor belt.

[0004] To achieve the above object, in the first aspect, the present invention provides a method for real-time monitoring of the deviation degree of a conveyor belt, including: Obtaining a historical image set of the conveyor belt running normally and deviating; wherein, the camera for obtaining images is installed directly above the axis of the conveyor belt; Marking the left and right idlers closest to the camera in each historical image of the historical image set to obtain a tagged image set; Performing enhancement processing on the tagged image set to obtain a training data set; Training YOLOv11 with the training data set to obtain a target segmentation model; Obtaining a real-time image of the conveyor belt running, and inputting the real-time image into the target segmentation model to obtain a segmentation feature image of the left and right idlers closest to the camera; Calculate the real-time conveyor belt deviation amount according to the segmentation feature images of the left and right idlers closest to the camera and a preset formula.

[0005] Optionally, perform enhancement processing on the labeled image set to obtain a training data set, including: Select any number of images from the labeled image set for copying and randomly add noise; Select any number of images from the labeled image set and the images with added noise for copying and randomly add occlusions; Select any number of images from the labeled image set, the images with added noise, and the images with added occlusions for copying and randomly adjust parameters; where the parameters include color, brightness, and contrast; Mix the labeled image set, the images with added noise, the images with added occlusions, and the images with adjusted parameters evenly to obtain a training data set.

[0006] Optionally, randomly adding occlusions includes: randomly adding one or more of shadows caused by light sources in different directions, the contour shapes of items on the conveyor belt, and light spots of different shapes and sizes.

[0007] Optionally, calculate the real-time conveyor belt deviation amount according to the segmentation feature images of the left and right idlers closest to the camera and a preset formula, including: Calculate the real-time pixel areas of the left and right idlers closest to the camera in the image according to the segmentation feature images of the left and right idlers closest to the camera; Based on the real-time pixel areas of the left and right idlers closest to the camera in the image and a preset formula, calculate the real-time conveyor belt deviation amount; where the preset formula is:

[0008] In the above formula, represents the deviation amount; represents the real-time pixel area of the left idler closest to the camera in the image; represents the real-time pixel area of the right idler closest to the camera in the image.

[0009] Optionally, calculate the real-time pixel areas of the left and right idlers closest to the camera in the image according to the segmentation feature images of the left and right idlers closest to the camera, including: Perform gray-scale conversion and filtering to reduce noise on the segmentation feature images of the left and right idlers closest to the camera to obtain processed segmentation feature images; Set different pixel values for the target area and the non-target area in the processed segmentation feature images; where the target area is the area where the left and right idlers closest to the camera are located; By counting the number of pixels in the target area and the total number of pixels in the segmented feature image, the proportion of the target area in the segmented feature image is obtained to obtain the real-time pixel area of the left and right idlers closest to the camera in the image.

[0010] Optionally, the method further includes: Obtain the pixel area of the idler on the left or right closest to the camera when the conveyor belt is not running off track as the reference area; Determine the warning threshold and the stop threshold according to the reference area; wherein, the warning threshold is less than the stop threshold; When the real-time conveyor belt deviation amount is greater than the warning threshold, perform deviation correction on the conveyor belt; When the real-time conveyor belt deviation amount is greater than the stop threshold, stop the conveyor belt.

[0011] Optionally, performing deviation correction on the conveyor belt includes: Compare the segmented feature images of the left and right idlers closest to the camera to obtain the deviation direction of the conveyor belt; Control the servo motor to adjust the idler bracket at a set angle according to the deviation amount, so that the rotation of the idler bracket is opposite to the deviation direction of the conveyor belt; Control the self-aligning idler to guide the conveyor belt to the center position through friction according to the real-time conveyor belt deviation amount to reduce the deviation amount; Continuously execute the step of reducing the deviation amount until the real-time conveyor belt deviation amount is less than the warning threshold, and stop the deviation correction.

[0012] In a second aspect, the present invention also provides a real-time conveyor belt deviation degree monitoring device, including: A historical image acquisition unit for acquiring a historical image set of the normal operation and deviation operation of the conveyor belt; wherein, the camera for acquiring images is installed directly above the axis of the conveyor belt; A historical image marking unit for marking the left and right idlers closest to the camera in each historical image of the historical image set to obtain a labeled image set; An image processing unit for enhancing the labeled image set to obtain a training data set; A model training unit for training YOLOv11 through the training data set to obtain a target segmentation model; A real-time image acquisition unit for acquiring a real-time image of the conveyor belt operation, inputting the real-time image into the target segmentation model, and obtaining a segmented feature image of the left and right idlers closest to the camera; A deviation amount calculation unit for calculating the real-time conveyor belt deviation amount according to the segmented feature images of the left and right idlers closest to the camera and a preset formula.

[0013] In a third aspect, the present invention also provides a machine-readable storage medium, on which instructions are stored for causing a machine to execute any one of the above real-time conveyor belt deviation degree monitoring methods.

[0014] In a fourth aspect, the present invention also provides a processor for running a program, wherein when the program is run, it is used to execute any one of the above real-time conveyor belt deviation degree monitoring methods.

[0015] The present invention accurately identifies the left and right idlers closest to each other in the image through a semantic segmentation model, realizes the accurate calculation of the deviation amount, thereby monitoring the deviation degree of the conveyor belt in real time, without the need to preset the edge points of the conveyor belt or identify the edge of the conveyor belt, with high recognition efficiency, reducing manual intervention, improving work efficiency and safety, and being applicable to the monitoring of conveyor belts in various industrial environments.

[0016] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent specific implementation part. Description of the Drawings

[0017] The drawings are used to provide a further understanding of the embodiments of the present invention, and constitute a part of the specification, and are used to explain the embodiments of the present invention together with the following specific implementation manners, but do not constitute a limitation to the embodiments of the present invention. In the drawings: Figure 1 is a flowchart of the real-time conveyor belt deviation degree monitoring method proposed by the present invention; Figure 2 is a schematic diagram of the camera position in an embodiment of the present invention; Figure 3 is the original image obtained by the camera in an embodiment of the present invention; Figure 4 is the image after enhancement processing in an embodiment of the present invention; Figure 5 is the segmentation feature image of the left and right idlers on the two sides closest to the camera in an embodiment of the present invention; Figure 6 is a schematic diagram of the calculation results of the real-time pixel area, conveyor belt deviation direction and deviation amount of the left and right idlers on the two sides closest to the camera in an embodiment of the present invention. Detailed Description of the Invention

[0018] The following will describe in detail the specific implementation manners of the embodiments of the present invention with reference to the drawings. It should be understood that the specific implementation manners described herein are only used to illustrate and explain the embodiments of the present invention, and are not used to limit the embodiments of the present invention.

[0019] It should be noted that the acquisition, transmission, storage, use, processing, etc. of data in the technical solution of this application all comply with relevant regulations. In the embodiments of this application, some existing solutions in the industry such as certain software, components, models, etc. may be mentioned. They should be regarded as exemplary, and their purpose is only to illustrate the feasibility in the implementation of the technical solution of this application, but it does not mean that the applicant has already or necessarily used this solution.

[0020] As Figure 1 shown, this embodiment provides a method for monitoring the deviation degree of a conveyor belt in real time, including: Obtain the historical image sets of the conveyor belt during normal operation and deviation operation; among them, the camera for obtaining images is installed directly above the axis of the conveyor belt.

[0021] Specifically, the above steps are used to obtain the performance of the conveyor belt in different states, providing a rich image basis for subsequent analysis and modeling. As Figure 2 shown, the camera is installed directly above the axis of the conveyor belt, so that the conveyor belt during normal operation and the idlers on the left and right sides of the conveyor belt captured by the camera are symmetrical objects, and the key features (such as idlers) in the image are easier to be identified and labeled. The historical image set and real-time image of the conveyor belt can be obtained by setting the camera to take pictures at fixed time intervals, or by video-recording the conveyor belt and intercepting video frames from the video at fixed time intervals.

[0022] Label the left and right idlers closest to the camera in each historical image of the historical image set to obtain a labeled image set.

[0023] Specifically, in this embodiment, the labelme software is used to label the left and right idlers closest to the camera. "left" is used to represent the left idler closest to the camera, and "right" is used to represent the right idler closest to the camera. The specific steps for labeling with the labelme software are: frame the left and right idlers closest to the camera in the image with a target box, and at the same time mark the positions of the idlers in the figure with a feature mask, which are divided into two categories: "left and right".

[0024] Perform enhancement processing on the labeled image set to obtain a training data set.

[0025] Train YOLOv11 with the training data set to obtain a target segmentation model.

[0026] Obtain the real-time image of the conveyor belt running, and input the real-time image into the target segmentation model to obtain the segmentation feature images of the left and right idlers closest to the camera.

[0027] Specifically, the results output by the target segmentation model include three results: a target box, a category, and a feature mask.

[0028] Based on the segmentation feature images of the left and right idlers closest to the camera and a preset formula, the real-time conveyor belt deviation amount is calculated.

[0029] Through the above steps, the image of the conveyor belt operation can be obtained in real time, quickly processed, and the deviation of the conveyor belt can be detected in time, avoiding production accidents or equipment damage caused by the deviation not being detected for a long time. Compared with the traditional manual inspection method, this method reduces the labor cost and can work continuously and efficiently, reducing the overall operation cost.

[0030] In some embodiments, the labeled image set is enhanced to obtain a training data set, including: Select any number of images from the labeled image set for copying and randomly add noise; Select any number of images from the labeled image set and the images with added noise for copying and randomly add occlusions; Select any number of images from the labeled image set, the images with added noise, and the images with added occlusions for copying and randomly adjust parameters; wherein, the parameters include color, brightness, and contrast; Mix the labeled image set, the images with added noise, the images with added occlusions, and the images with adjusted parameters evenly to obtain a training data set.

[0031] Specifically, the randomly added noise includes one or more of Gaussian noise, salt-and-pepper noise, and Poisson noise. Among them, Gaussian noise is used to simulate the noise generated by the camera; salt-and-pepper noise is used to simulate the interference suffered by the image during transmission; Poisson noise is used to simulate statistical noise. By randomly adjusting the color deviation of the image, the diversity of the training data set is increased. By randomly changing the brightness of the image, it is used to simulate the change of illumination. By adjusting the contrast of the image, the light and dark difference in the image is made more obvious or more blurred, which is used to simulate working environments with different visibility.

[0032] In some embodiments, randomly adding occlusions includes: randomly adding one or more of shadows caused by light sources in different directions, the contour shapes of objects on the conveyor belt, and light spots of different shapes and sizes.

[0033] Specifically, the shadows caused by light sources in different directions include the shadows formed by the conveyor belt, the objects on the conveyor belt, and the objects in the actual application environment under the illumination of light sources in different directions, which are used to simulate the possible shadow situations in the actual application environment. The contour shapes of objects on the conveyor belt are used to simulate the situation where part of the camera's field of view may be occluded in the actual application environment. Light spots of different shapes and sizes are used to simulate the imaging of sunlight, light, or reflected light in the camera in the actual application environment.

[0034] In some embodiments, any number of images after being uniformly mixed can be selected for copying and adding dynamic blur to simulate the real effect when the conveyor belt runs at high speed.

[0035] The purpose of the above image enhancement processing is to expand the training data set, make the images closer to various real environments and extreme environments, and effectively improve the accuracy and generalization ability of the target segmentation model after training. Figure 3 is the original image captured by the camera, Figure 4 is the image after the original image enhancement processing, through Figure 3 and Figure 4 By comparison, it can be seen that it is almost impossible to see the position and features of the idler in the picture with the naked eye for the photo after the enhancement processing, which well simulates the running condition of the conveyor belt in extreme environments (such as environments with occlusion, severe coal dust pollution, camera lens pollution, etc.).

[0036] In some embodiments, according to the segmentation feature images of the left and right idlers closest to the camera and a preset formula, the real-time conveyor belt deviation amount is calculated, including: According to the segmentation feature images of the left and right idlers closest to the camera, calculate the real-time pixel areas of the left and right idlers closest to the camera in the image; Based on the real-time pixel areas of the left and right idlers closest to the camera in the image and a preset formula, calculate the real-time conveyor belt deviation amount; wherein, the preset formula is:

[0037] In the above formula, represents the deviation amount; represents the real-time pixel area of the left idler closest to the camera in the image; represents the real-time pixel area of the right idler closest to the camera in the image.

[0038] Specifically, by calculating the real-time pixel areas of the left and right idlers closest to the camera in the image, the position difference of the idlers can be accurately quantified. This is because the camera is located on the center line of the two idlers. When the conveyor belt does not deviate, the areas of the left and right idlers should be the same. Therefore, when the areas of the left and right idlers are different, it means that the conveyor belt deviates. Selecting the idlers closest to the camera, one reason is that the closer to the camera, the clearer the shooting, which is beneficial to clearly identify the edge of the idler. The other reason is that the closer to the camera, the larger the area occupied by the idler in the image, so as to more accurately obtain the deviation degree of the conveyor belt. The present invention uses a simple pixel area ratio to calculate the deviation amount, reduces the complexity of the algorithm, improves the calculation efficiency, and is convenient for real-time processing of a large amount of image data. In this embodiment, the segmentation feature images of the left and right idlers closest to the camera are as Figure 5As shown in the figure, "left" in the figure represents the left idler closest to the camera, and "right" represents the right idler closest to the camera.

[0039] In some embodiments, based on the segmented feature images of the left and right idlers closest to the camera, the real-time pixel areas of the left and right idlers closest to the camera in the image are calculated, including: Perform gray-scale conversion and noise filtering on the segmented feature images of the left and right idlers closest to the camera to obtain the processed segmented feature images; Set different pixel values for the target area and the non-target area in the processed segmented feature images; wherein, the target area is the area where the left and right idlers closest to the camera are located; By counting the number of pixel points in the target area and the total number of pixel points in the segmented feature image, the proportion of the target area in the segmented feature image is obtained to obtain the real-time pixel areas of the left and right idlers closest to the camera in the image.

[0040] In some embodiments, after calculating the real-time conveyor belt deviation amount, it further includes: Obtain the pixel area of the idler on the left or right closest to the camera when the conveyor belt is not deviated as the reference area; Determine the warning threshold and the stop threshold according to the reference area; wherein, the warning threshold is less than the stop threshold; When the real-time conveyor belt deviation amount is greater than the warning threshold, perform deviation correction on the conveyor belt; When the real-time conveyor belt deviation amount is greater than the stop threshold, stop the conveyor belt.

[0041] Specifically, by setting the warning threshold and the stop threshold, an alarm can be issued in advance when the conveyor belt deviation amount reaches a certain level, and deviation correction or stop processing can be performed. The setting of the warning threshold and the stop threshold realizes a phased processing strategy. The warning threshold is used for early intervention, and the stop threshold is used for emergency stop, which can not only ensure production efficiency but also ensure safety. Usually, the warning threshold is set to 10% of the reference area, and the stop threshold is set to 20% of the reference area. The warning threshold and the stop threshold can be flexibly adjusted according to the actual operating conditions and requirements to adapt to different working conditions and application scenarios, increasing the adaptability and flexibility of the system.

[0042] In some embodiments, performing deviation correction on the conveyor belt includes: Compare the segmented feature images of the left and right idlers closest to the camera to obtain the deviation direction of the conveyor belt; Control the servo motor to adjust the idler bracket at a set angle according to the deviation amount, so that the rotation of the idler bracket is opposite to the deviation direction of the conveyor belt; According to the real-time deviation of the conveyor belt, control the self-aligning idler to guide the conveyor belt to the center position through friction to reduce the deviation; Continuously execute the steps to reduce the deviation until the real-time deviation of the conveyor belt is less than the warning threshold, and then stop the deviation correction.

[0043] Specifically, compare the segmented feature images of the left and right idlers closest to the camera. If the pixel area of the left idler is larger than that of the right idler, it means that the conveyor belt is running to the right and needs to be adjusted to the left. Since there is a certain lag in the deviation correction process, it is necessary to stop the deviation correction when the real-time deviation of the conveyor belt is less than the warning threshold. If the stop deviation correction signal is sent when the deviation is 0, new deviation will be caused due to the lag. Through real-time monitoring and automatic deviation correction, production interruptions and failures caused by conveyor belt deviation are reduced, and production efficiency and equipment utilization rate are improved. Through automatic deviation correction, the stability and reliability of the system are enhanced, and the errors and failure risks caused by improper manual operation are reduced. In this embodiment, Siemens S7-200 PLC is selected to control the servo motor and the self-aligning idler. The real-time pixel areas of the left and right idlers closest to the camera calculated in this embodiment and the calculation results of the deviation amount are as Figure 6 shown. In the figure, "left: 4158.00" represents the real-time pixel area of the left idler closest to the camera in the image is 4158.00; "right: 4890.00" represents the real-time pixel area of the right idler closest to the camera in the image is 4890.00; "leftdeviation" represents that the conveyor belt is running to the left; "D: 0.08" represents that the deviation amount is 0.08.

[0044] This embodiment also provides a device for monitoring the deviation degree of the real-time conveyor belt, including: A historical image acquisition unit, which is used to acquire a set of historical images of the conveyor belt running normally and deviating; among them, the camera used to acquire images is installed directly above the axis of the conveyor belt; A historical image marking unit, which is used to mark the left and right idlers closest to the camera in each historical image of the historical image set to obtain a set of labeled images; An image processing unit, which is used to perform enhancement processing on the set of labeled images to obtain a training data set; A model training unit, which is used to train YOLOv11 through the training data set to obtain a target segmentation model; A real-time image acquisition unit, which is used to acquire a real-time image of the conveyor belt running, input the real-time image into the target segmentation model, and obtain the segmented feature images of the left and right idlers closest to the camera; A deviation amount calculation unit, which is used to calculate the real-time deviation of the conveyor belt according to the segmented feature images of the left and right idlers closest to the camera and a preset formula.

[0045] The real-time conveyor belt deviation degree monitoring device includes a processor and a memory. The above-mentioned historical image acquisition unit, historical image marking unit, image processing unit, model training unit, real-time image acquisition unit, deviation amount calculation unit, and conveyor belt adjustment unit, etc., are all stored in the memory as program units, and the processor executes the above program units stored in the memory to implement corresponding functions.

[0046] The processor contains a kernel, and the kernel retrieves the corresponding program units from the memory. One or more kernels can be set, and the real-time monitoring of the conveyor belt deviation degree is achieved by adjusting the kernel parameters.

[0047] The memory may include non-permanent memory in a computer-readable medium, forms such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one storage chip.

[0048] An embodiment of the present invention provides a storage medium, on which a program is stored, and when the program is executed by a processor, it implements the real-time conveyor belt deviation degree monitoring method.

[0049] An embodiment of the present invention provides a processor, and the processor is used to run a program, wherein when the program runs, it executes the real-time conveyor belt deviation degree monitoring method.

[0050] An embodiment of the present invention provides a device, which includes a processor, a memory, and a program stored on the memory and executable on the processor. The device herein can be a server, a PC, a PAD, a mobile phone, etc.

[0051] This application also provides a computer program product, which is suitable for executing a program for initializing the steps of the real-time conveyor belt deviation degree monitoring method when executed on a data processing device.

[0052] Those skilled in the art should understand that the embodiments of the present application can provide a method, a system, or a computer program product. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0053] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, as well as the combination of flows and / or blocks in the flowchart and / or block diagram. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing device produce a means for implementing the functions specified in the flow Figure 1 one flow or multiple flows and / or blocks Figure 1 means for implementing the functions specified in one block or multiple blocks.

[0054] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including an instruction means that implements the functions specified in the flow Figure 1 one flow or multiple flows and / or blocks Figure 1 means for implementing the functions specified in one block or multiple blocks.

[0055] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in the flow Figure 1 one flow or multiple flows and / or blocks Figure 1 means for implementing the functions specified in one block or multiple blocks.

[0056] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and a memory.

[0057] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM), and / or non-volatile memory such as read-only memory (ROM) or flash memory (flash RAM). The memory is an example of computer-readable media.

[0058] A computer-readable medium includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory, or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD), or other optical storage, magnetic cassette tapes, magnetic disk storage, or other magnetic storage devices, or any other non-transitory medium that can be used to store information accessible by a computing device. As defined herein, a computer-readable medium does not include transitory computer-readable media, such as modulated data signals and carrier waves.

[0059] A real-time conveyor belt deviation degree monitoring method, device, storage medium, and processor based on YOLOv11 semantic segmentation proposed by the present invention abandon the past complex deep learning models and conveyor belt detection mathematical models, use the preset formula proposed by the present invention, and combine the characteristics of the symmetric structure of the conveyor belt system to realize the monitoring of conveyor belt deviation by segmenting the left and right idlers closest to the camera. The invention simplifies complexity, converts the complex real-time monitoring task into the monitoring of the pixel area of the idlers, has strong practicability, real-time performance, and accuracy, and can accurately complete the conveyor belt deviation monitoring task even in extreme environments. By setting warning thresholds and shutdown thresholds, it is also possible to perform real-time correction on conveyor belts with minor deviations and perform safe shutdown on conveyor belts with severe deviations.

[0060] It should also be noted that the terms "include", "comprise", or any other variant thereof are intended to cover non-exclusive inclusion, such that a process, method, commodity, or device including a series of elements not only includes those elements but also includes other elements not expressly listed, or further includes elements inherent to such process, method, commodity, or device. Without further limitation, an element defined by the statement "including one" does not exclude the existence of additional identical elements in the process, method, commodity, or device including the element.

[0061] The above are only examples of the present application and are not used to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.

Claims

1. A real-time conveyor belt deviation monitoring method, characterized in that: include: Obtain a historical image set of the normal operation and deviation operation of the conveyor belt; wherein the camera used to obtain the image is installed directly above the axis of the conveyor belt; Mark the left and right rollers closest to the camera in each historical image of the historical image set to obtain a labeled image set; Perform enhancement processing on the labeled image set to obtain a training data set; Train YOLOv11 with the training data set to obtain the target segmentation model; Acquire the real-time image of the conveyor belt running, input the real-time image into the target segmentation model, and obtain the segmentation feature image of the left and right rollers closest to the camera; The real-time conveyor belt deviation is calculated based on the segmented feature images of the left and right rollers closest to the camera and the preset formula.

2. The real-time conveyor belt deviation monitoring method according to claim 1 is characterized in that: The labeled image set is enhanced to obtain a training data set, including: Select any number of images from the labeled image set, replicate them and randomly add noise; Select any number of images from the labeled image set and the image with added noise to replicate and randomly add occlusions; Select any number of images from the labeled image set, the image after adding noise, and the image after adding occlusion, copy them, and randomly adjust the parameters; wherein the parameters include color, brightness, and contrast; The labeled image set, the image with added noise, the image with added occlusion, and the image with adjusted parameters are mixed evenly to obtain the training data set.

3. The real-time conveyor belt deviation monitoring method according to claim 2 is characterized in that: Randomly adding occlusions includes: randomly adding one or more of shadows caused by light sources in different directions, the outline shapes of objects on the conveyor belt, and light spots of different shapes and sizes.

4. The real-time conveyor belt deviation monitoring method according to claim 1 is characterized in that: According to the segmented feature images of the left and right rollers closest to the camera and the preset formula, the real-time conveyor belt deviation is calculated, including: According to the segmented feature images of the left and right rollers closest to the camera, the real-time pixel areas of the left and right rollers closest to the camera in the image are calculated; The real-time conveyor belt deviation is calculated based on the real-time pixel area of ​​the left and right rollers closest to the camera in the image and a preset formula; wherein the preset formula is: In the above formula, Indicates the deviation amount; Indicates the real-time pixel area of ​​the left roller closest to the camera in the image; Represents the real-time pixel area of ​​the right roller closest to the camera in the image.

5. The real-time conveyor belt deviation monitoring method according to claim 4 is characterized in that: According to the segmented feature images of the left and right rollers closest to the camera, the real-time pixel areas of the left and right rollers closest to the camera in the image are calculated, including: Perform grayscale conversion and filtering noise reduction on the segmentation feature images of the left and right rollers closest to the camera to obtain a processed segmentation feature image; Different pixel values ​​are set for the target area and the non-target area in the processed segmented feature image; wherein the target area is the area where the left and right rollers closest to the camera are located; By counting the number of pixels in the target area and the total number of pixels in the segmented feature image, the proportion of the target area in the segmented feature image is obtained, so as to obtain the real-time pixel area of ​​the left and right rollers closest to the camera in the image.

6. The real-time conveyor belt deviation monitoring method according to claim 1 is characterized in that: The method further comprises: The pixel area of ​​the left or right roller closest to the camera when the conveyor belt is not deviating in the image is obtained as the reference area; According to the reference area, a warning threshold and a shutdown threshold are determined; wherein the warning threshold is smaller than the shutdown threshold; When the real-time conveyor belt deviation is greater than the warning threshold, the conveyor belt is corrected; When the real-time conveyor belt deviation is greater than the shutdown threshold, the conveyor belt is shut down.

7. The real-time conveyor belt deviation monitoring method according to claim 6 is characterized in that: Correction of conveyor belt, including: Compare the segmented feature images of the left and right rollers closest to the camera to obtain the deviation direction of the conveyor belt; According to the deviation amount, the servo motor is controlled to adjust the roller frame according to the set angle, so that the rotation of the roller frame is opposite to the deviation direction of the conveyor belt; According to the real-time conveyor belt deviation, the self-aligning roller is controlled to guide the conveyor belt to the center position through friction to reduce the deviation; Continue to execute the steps to reduce the deviation amount until the real-time conveyor belt deviation amount is less than the warning threshold, and stop correcting the deviation.

8. A real-time conveyor belt deviation monitoring device, characterized in that: include: A historical image acquisition unit is used to acquire a historical image set of the normal operation and deviation operation of the conveyor belt; wherein the camera used to acquire the image is installed directly above the axis of the conveyor belt; A historical image marking unit is used to mark the left and right rollers closest to the camera in each historical image of the historical image set to obtain a labeled image set; An image processing unit, used for performing enhancement processing on the labeled image set to obtain a training data set; A model training unit is used to train YOLOv11 using a training data set to obtain a target segmentation model; A real-time image acquisition unit is used to acquire a real-time image of the conveyor belt running, input the real-time image into the target segmentation model, and obtain the segmentation feature images of the left and right rollers closest to the camera; The deviation calculation unit is used to calculate the real-time conveyor belt deviation according to the segmented feature images of the left and right rollers closest to the camera and a preset formula.

9. A machine-readable storage medium, characterized in that: The machine-readable storage medium stores instructions for enabling the machine to execute the real-time conveyor belt deviation degree monitoring method as described in any one of claims 1-7.

10. A processor, characterized in that: Used to run a program, wherein the program, when run, is used to execute the real-time conveyor belt deviation degree monitoring method described in any one of claims 1-7.

Images