Belt conveyor tearing defect detection method and system based on machine vision

Abstract

The invention provides a belt conveyor tearing defect detection method and system based on machine vision. Laser stripes are formed on the surface of the back face of a belt; laser stripe distortion images generated by belt tearing of the belt conveyor under various working conditions and laser stripe images under normal working conditions are obtained; performing preprocessing and data enhancement on the obtained laser stripe image to form a laser stripe data set under various working conditions; dividing the laser stripe data set, and performing feature extraction and fusion on the laser stripe data set by using a deep neural network model to realize training of the deep neural network model; performing classification detection on the processed to-be-detected image by using the trained deep neural network model; according to a belt tearing defect classification detection result, an alarm decision factor is introduced, a judgment level corresponding to the defect is determined, and corresponding early warning is sent out. The tearing defect detection accuracy of the belt conveyor can be improved, a corresponding alarm decision scheme is given, and the production and maintenance efficiency is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of coal mine automation online monitoring, and in particular relates to a belt conveyor tear defect detection method and system based on machine vision. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] In the coal mining industry, belt conveyors are one of the core equipment for material transportation, responsible for transporting mined coal from underground to storage or processing facilities on the ground. These conveying systems usually carry a large amount of coal, so their performance and reliability are extremely high. However, a major problem that belt conveyors often encounter during long-term operation is belt tearing.

[0004] Belt tearing may be caused by a variety of factors, such as impact from sharp objects in the material, failure of conveyor accessories (such as wear of guide wheels or rollers), belt deviation tearing and core pull belt tearing. Belt tearing will not only cause coal leakage and increase material loss, but may also cause the entire conveyor line to shut down, seriously affecting the production efficiency of the coal mine. In addition, belts with excessive tearing may bring greater safety hazards and increase the inspection and maintenance costs of the system.

[0005] In recent years, with the increasing application of machine vision technology in the industrial field, the use of machine vision technology instead of traditional belt tear detection technology can effectively improve production efficiency and product quality. However, the alarm decision-making mechanism of existing machine vision detection methods is not perfect. Often, when a tear defect is detected, an alarm is issued and the machine is shut down, which will cause additional losses to the enterprise. Summary of the invention

[0006] In order to solve the above problems, the present invention proposes a belt conveyor tear defect detection method and system based on machine vision. The present invention can improve the accuracy of belt conveyor tear defect detection, and provide corresponding alarm decision-making solutions to improve production and maintenance efficiency.

[0007] According to some embodiments, the present invention adopts the following technical solutions:

[0008] A belt conveyor tear defect detection method based on machine vision comprises the following steps:

[0009] Forming laser stripes on the surface of the back of the belt;

[0010] Obtain laser stripe distortion images caused by belt tearing under various working conditions of the belt conveyor and laser stripe images under normal working conditions;

[0011] Preprocess and enhance the acquired laser stripe images to form laser stripe data sets under various working conditions;

[0012] Dividing the laser stripe data set, extracting and fusing features of the data set using a deep neural network model, and training the deep neural network model;

[0013] Obtain the image to be detected, and perform preprocessing and data enhancement on the image to be detected;

[0014] Use the trained deep neural network model to classify and detect the processed images to be detected;

[0015] According to the classification detection results of belt tear defects, the alarm decision factor is introduced to determine the judgment level corresponding to the defect and issue the corresponding warning.

[0016] As an optional embodiment, a line laser transmitter is used to emit multiple lines of laser to the back side of the belt to form laser stripes on the surface of the back side of the belt;

[0017] An industrial camera is used to collect laser stripe distortion images caused by belt tearing under various working conditions of the belt conveyor and laser stripe images under normal working conditions.

[0018] As an optional implementation, the preprocessing process includes smoothing the image using Gaussian filtering, grayscale processing the image, converting the color image into a grayscale image with only black and white colors, and corroding the grayscale image to reduce noise in the image.

[0019] As an optional implementation, the data enhancement operation includes: randomly stretching, rotating, scaling, translating, cropping and brightness conversion of the pre-processed image.

[0020] As an optional implementation, the process of forming a laser stripe data set under various working conditions includes labeling the image data that has completed the data enhancement operation according to the four characteristics of the belt tear defect, namely, depth, length, width and density.

[0021] As an optional implementation, the process of introducing the alarm decision factor includes:

[0022] The alarm decision function is set based on the depth, length, width and density characteristics of the belt tear defect in the labeled data:

[0023] Δ=ω d d+ω l l+ω w w+ω s s;

[0024] Among them, Δ is the alarm decision factor, ω dis the tear depth coefficient, ω l is the tear length coefficient, ω w is the tear width coefficient, ω s is the tear density coefficient, d is the tear depth, l is the tear length, w is the tear width, and s is the tear density.

[0025] As a further example, d ,ω l ,ω w ,ω s The specific value of is determined using a regression model in machine learning.

[0026] As an optional implementation method, the process of determining the judgment level corresponding to the defect and issuing the corresponding warning includes dividing the judgment level into normal, slight damage, heavier damage and severe damage according to the size of the alarm decision factor. When the judgment result is heavier damage, a warning is issued and the location distribution of the damage is reported. When the judgment result is severe damage, an alarm is immediately issued and the machine is shut down.

[0027] A belt conveyor tear defect detection system based on machine vision, comprising:

[0028] An image acquisition module is configured to form laser stripes on the surface of the back of the belt, acquire the laser stripe distortion images caused by belt tearing under various working conditions of the belt conveyor and the laser stripe images under normal working conditions, and acquire the image to be detected;

[0029] A data processing module is configured to preprocess and enhance the data of the acquired laser stripe image to form a laser stripe data set under various working conditions, and to preprocess and enhance the data of the image to be detected;

[0030] A model training module is configured to divide the laser stripe data set, extract and fuse features thereof using a deep neural network model, and implement training of the deep neural network model;

[0031] A classification detection module is configured to perform classification detection on the processed image to be detected using the trained deep neural network model;

[0032] The early warning module is configured to introduce alarm decision factors according to the belt tear defect classification detection results, determine the judgment level corresponding to the defect, and issue a corresponding early warning.

[0033] An electronic device comprises a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps in the above method are completed.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] The present invention can quickly identify belt tear defects. On the basis of retaining the advantages of belt tear detection technology based on machine vision, the present invention further improves the alarm decision mechanism of belt tear defect detection, avoids additional losses caused by frequent alarm shutdowns, and improves the efficiency of coal mine production.

[0036] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0038] Figure 1 is a flow chart of belt conveyor tear defect detection based on machine vision in one embodiment;

[0039] Figure 2 The present invention is a flowchart of pre-processing operation in the process of detecting tear defects of a belt conveyor based on machine vision according to an embodiment;

[0040] Figure 3 The present invention is a schematic diagram of an alarm decision mechanism in a belt conveyor tear defect detection process based on machine vision according to an embodiment;

[0041] Figure 4 This is a schematic diagram of the relative installation positions of various detection devices in an embodiment. DETAILED DESCRIPTION

[0042] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0043] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0044] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0045] In the absence of conflict, the embodiments in this application and the features in the embodiments may be combined with each other.

[0046] Embodiment 1

[0047] A belt conveyor tear defect detection method based on machine vision, while retaining the advantages of the belt tear detection technology based on machine vision, proposes a more complete alarm decision-making mechanism and a more practical belt tear defect positioning method. The specific steps are as follows: Figure 1 As shown:

[0048] S100: Use the line laser transmitter to emit a red line laser to the back of the belt, and the line laser will form a red laser stripe on the surface of the back of the belt;

[0049] S200: Use CCD industrial camera to collect laser stripe distortion images caused by belt tearing under various working conditions of belt conveyor and laser stripe images under normal working conditions;

[0050] Of course, in other embodiments, other methods may be used to form laser stripes, or to form laser images of other colors or types.

[0051] Likewise, in other embodiments, other devices may be used to capture images.

[0052] like Figure 4 As shown, in this embodiment, the line laser emitter and the CCD industrial camera are arranged side by side near the belt. However, in other embodiments, the laser emitting device and the image acquisition device may not be on the same side, as long as it is possible to form a laser image on the belt or obtain an image on the belt.

[0053] S300: pre-processing the collected laser stripe images to generate laser stripe data sets of the belt conveyor under various working conditions;

[0054] In this embodiment, the image preprocessing step is as follows: Figure 2 As shown, including:

[0055] S310: using Gaussian filtering to smooth the image to reduce noise and other unnecessary pixel information (such as dust generated during coal mine transportation), thereby improving image quality and clarity;

[0056] S320: grayscale processing is performed on the image, and a weighted average method is used to convert the color image originally having three color channels of R, G, and B into a grayscale image with only a single channel, thereby reducing the amount of data in the image, reducing memory usage, and improving the computing speed;

[0057] S330: performing an erosion operation on the grayscale image to further reduce noise and small objects that may exist in the image;

[0058] S340: performing a data enhancement operation on the image after the erosion operation is performed to improve the generalization ability of the model;

[0059] In this embodiment, the data enhancement operation includes randomly stretching, rotating, scaling, translating, cropping and brightness conversion operations on the laser stripe image;

[0060] S350: annotating the image data obtained in the above process according to the four characteristics of the belt tear defect, namely, depth, length, width, and density;

[0061] S400: Divide the data set, use deep neural network to do deep learning on image data, perform feature extraction and fusion, and use it for image recognition and classification tasks;

[0062] In this embodiment, the deep neural network model is YOLOv10, which outputs the predicted values ​​of the depth, length, width and density of the belt tear after model prediction, and marks the location where the tear occurs with a rectangular box;

[0063] Of course, in other embodiments, other deep neural network models, such as R-CNN, etc., may be used.

[0064] S500: Deploy YOLOv10 on NVIDIA GPU RTX4060 for image recognition and classification detection;

[0065] It should be noted that when real-time belt tear defect detection is performed using a CCD industrial camera, the images to be detected are processed in the same way as the images used for training;

[0066] S600: The detection work is continuously carried out. When a belt tear is detected, an early warning is issued according to a pre-set alarm decision mechanism, and the location distribution of the belt tear is reported;

[0067] In this embodiment, the alarm decision mechanism is specifically:

[0068] like Figure 3 As shown in the figure, the alarm decision function is set based on the four features of the depth, length, width and density of the belt tear defects in the labeled data:

[0069] Δ=ω d d+ω l l+ω w w+ω s s;

[0070] Where: Δ is the alarm decision factor, ω d is the tear depth coefficient, ω l is the tear length coefficient, ω w is the tear width coefficient, ωs is the tear density coefficient, d is the tear depth, l is the tear length, w is the tear width, and s is the tear density;

[0071] In this embodiment, ω d ,ω l ,ω w ,ω s The specific value of is determined using the multivariate regression model in machine learning, and once the value is determined, it will not be changed again;

[0072] For those in this field, the training process of the multivariate regression model is common knowledge, so the following steps are only briefly explained:

[0073] Use the same dataset as the deep learning task and divide the dataset into 60% for training, 20% for cross-validation, and 20% for prediction.

[0074] Determine the multivariate regression model and loss function, use gradient descent to update the weights, and determine the specific value of the weights when the loss function converges and the loss is less than the set threshold;

[0075] The weight is ω d ,ω l ,ω w ,ω s Specific value of

[0076] In this embodiment, the specific values ​​of d, l, w, and s are predicted by the deep neural network YOLOv10.

[0077] According to the size of Δ, the judgment level is divided into four categories: normal, slight damage, relatively serious damage, and severe damage. When the judgment result is relatively serious damage, an early warning is issued and the location distribution of the damage is reported. When the judgment result is severe damage, an alarm is immediately issued and the machine is shut down.

[0078] Embodiment 2

[0079] A belt conveyor tear defect detection system based on machine vision, comprising:

[0080] An image acquisition module is configured to form laser stripes on the surface of the back of the belt, acquire the laser stripe distortion images caused by belt tearing under various working conditions of the belt conveyor and the laser stripe images under normal working conditions, and acquire the image to be detected;

[0081] A data processing module is configured to preprocess and enhance the data of the acquired laser stripe image to form a laser stripe data set under various working conditions, and to preprocess and enhance the data of the image to be detected;

[0082] A model training module is configured to divide the laser stripe data set, extract and fuse features thereof using a deep neural network model, and implement training of the deep neural network model;

[0083] A classification detection module is configured to perform classification detection on the processed image to be detected using the trained deep neural network model;

[0084] The early warning module is configured to introduce alarm decision factors according to the belt tear defect classification detection results, determine the judgment level corresponding to the defect, and issue a corresponding early warning.

[0085] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take 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.

[0086] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0087] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0088] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0089] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principle of the present invention without creative labor shall be included in the protection scope of the present invention.

Claims

1. A belt conveyor tear defect detection method based on machine vision, characterized in that: The following steps are involved: Forming laser stripes on the surface of the back of the belt; Obtain laser stripe distortion images caused by belt tearing under various working conditions of the belt conveyor and laser stripe images under normal working conditions; Preprocess and enhance the acquired laser stripe images to form laser stripe data sets under various working conditions; Dividing the laser stripe data set, extracting and fusing features of the data set using a deep neural network model, and training the deep neural network model; Obtain the image to be detected, and perform preprocessing and data enhancement on the image to be detected; Use the trained deep neural network model to classify and detect the processed images to be detected; According to the classification detection results of belt tear defects, the alarm decision factor is introduced to determine the judgment level corresponding to the defect and issue the corresponding warning.

2. A belt conveyor tear defect detection method based on machine vision as claimed in claim 1, characterized in that: Use a line laser transmitter to emit multiple line lasers to the back of the belt to form laser stripes on the surface of the back of the belt; An industrial camera is used to collect laser stripe distortion images caused by belt tearing under various working conditions of the belt conveyor and laser stripe images under normal working conditions.

3. A belt conveyor tear defect detection method based on machine vision as claimed in claim 1, characterized in that: The preprocessing process includes using Gaussian filtering to smooth the image, grayscale processing the image, converting the color image into a grayscale image with only black and white colors, and corroding the grayscale image to reduce the noise in the image.

4. A belt conveyor tear defect detection method based on machine vision as claimed in claim 1, characterized in that: The data enhancement operation includes: randomly stretching, rotating, scaling, translating, cropping and brightness conversion of the pre-processed image.

5. The machine vision-based belt conveyor tear defect detection method according to claim 1, characterized in that: The process of forming a laser stripe data set under various working conditions includes labeling the image data that has completed the data enhancement operation according to the four characteristics of the belt tear defect, namely, depth, length, width and density.

6. A belt conveyor tear defect detection method based on machine vision as claimed in claim 1, characterized in that: The process of alarm decision factor includes: The alarm decision function is set based on the depth, length, width and density characteristics of the belt tear defect in the labeled data: D=oh d d+ω l l+ω w w+ω s s; Among them, Δ is the alarm decision factor, ω d is the tear depth coefficient, ω l is the tear length coefficient, ω w is the tear width coefficient, ω s is the tear density coefficient, d is the tear depth, l is the tear length, ω is the tear width, and s is the tear density.

7. A belt conveyor tear defect detection method based on machine vision as claimed in claim 6, characterized in that: ω d ,ω l ,ω w ,ω s The specific value of is determined using a regression model in machine learning.

8. The machine vision-based belt conveyor tear defect detection method according to claim 1, characterized in that: The process of determining the judgment level corresponding to the defect and issuing the corresponding warning includes dividing the judgment level into normal, slight damage, relatively serious damage and severe damage according to the size of the alarm decision factor. When the judgment result is relatively serious damage, a warning is issued and the location distribution of the damage is reported. When the judgment result is severe damage, an alarm is immediately issued and the machine is shut down.

9. A belt conveyor tear defect detection system based on machine vision, characterized in that: include: An image acquisition module is configured to form laser stripes on the surface of the back of the belt, acquire the laser stripe distortion images caused by belt tearing under various working conditions of the belt conveyor and the laser stripe images under normal working conditions, and acquire the image to be detected; A data processing module is configured to preprocess and enhance the data of the acquired laser stripe image to form a laser stripe data set under various working conditions, and to preprocess and enhance the data of the image to be detected; A model training module is configured to divide the laser stripe data set, extract and fuse features thereof using a deep neural network model, and implement training of the deep neural network model; A classification detection module is configured to perform classification detection on the processed image to be detected using the trained deep neural network model; The early warning module is configured to introduce alarm decision factors according to the belt tear defect classification detection results, determine the judgment level corresponding to the defect, and issue a corresponding early warning.

10. An electronic device, characterized in that: The method comprises a memory and a processor, and computer instructions stored in the memory and executed on the processor, wherein when the computer instructions are executed by the processor, the steps in the method according to any one of claims 1 to 8 are completed.

Images