Conveyor belt tearing detection method and device based on YOLACT network
By employing a YOLACT network-based method on conveyor belts, combined with a forward-oriented detection method, the high precision and reliability of conveyor belt tear detection were improved, the false alarm rate and false negative rate were reduced, and accurate and rapid detection of conveyor belt tears was achieved.
Patent Information
- Application Number
- CN202511123820.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-12
- Publication Date
- 2025-12-12
AI Technical Summary
Traditional conveyor belt tear detection methods rely on manual inspection, which suffers from problems such as detection lag, high false alarm rate and poor adaptability. Furthermore, existing visual inspection methods lack accuracy and reliability in complex industrial environments.
A YOLACT network-based method is used to acquire images and extract the center lines of the stripes by emitting straight laser stripes onto the lower surface of the conveyor belt. A deep learning network is then used for tear detection, and median filtering and CLAHE processing are combined to improve image quality and trigger an alarm signal.
It achieves high precision and reliability in conveyor belt tear detection, reduces false alarm rate and false alarm rate, and can accurately and quickly detect tears in complex environments.
Smart Images
Figure CN121120505A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of conveyor belt detection, and particularly relates to a conveyor belt tearing detection method and device based on a YOLACT network. BACKGROUND
[0002] As a key equipment for transporting materials, the running state of a conveyor belt directly affects production efficiency and job safety. Traditional conveyor belt tearing detection methods mainly rely on manual inspection, and have problems such as detection lag, high false alarm rate and poor adaptability.
[0003] Manual inspection is limited by the experience of the inspection personnel, and it is difficult to achieve all-weather and high-precision real-time monitoring. When the conveyor belt is running at high speed, manual judgment is prone to errors.
[0004] With the development of computer vision and deep learning technology, non-contact detection methods based on machine vision have gradually become a research hotspot. However, existing visual detection methods mostly rely on single features for judgment, such as identifying tearing only through image gray level changes or edge features. In complex industrial environments, they are easily disturbed by changes in light and stains, resulting in insufficient detection accuracy and reliability. SUMMARY
[0005] Therefore, it is necessary to provide a conveyor belt tearing detection method and device based on a YOLACT network to improve the accuracy and real-time performance of conveyor belt tearing detection.
[0006] To achieve the above purpose, in a first aspect, the present application provides a conveyor belt tearing detection method based on a YOLACT network, comprising: emitting a straight laser stripe to the lower surface of a conveyor belt to be detected by using a laser, and acquiring a lower surface image of the conveyor belt to be detected; extracting a stripe center of the laser stripe in the lower surface image, and obtaining a stripe center line based on the stripe center; inputting the stripe center line into a trained detection model to obtain a detection result output by the detection model; the detection model is obtained by training a YOLACT network based on a sample stripe center line of a sample lower surface image and a sample label of the sample lower surface image; the sample label includes a tearing degree corresponding to the sample lower surface image.
[0007] In a possible implementation manner, the detection model is obtained by training in the following manner: labeling the tearing degree of the sample lower surface image to obtain a sample label; the sample label includes: no tearing, slight tearing and severe tearing; The YOLACT network is trained based on the center line of the sample stripes and the sample labels, with the minimum loss function as the learning objective, to obtain the detection model.
[0008] In one possible implementation, the loss function is expressed as follows:
[0009] in, Represents the loss function. Indicates the actual value. Indicates the predicted value. m This indicates the number of images of the lower surface of the sample.
[0010] In one possible implementation, after emitting straight laser stripes onto the lower surface of the conveyor belt to be inspected using a laser and acquiring an image of the lower surface of the conveyor belt, the method further includes: The lower surface image is subjected to median filtering and CLAHE contrast enhancement processing.
[0011] In one possible implementation, after inputting the stripe center line into the trained detection model and obtaining the detection result output by the detection model, the method further includes: When the detection results indicate that the conveyor belt under test has torn, an alarm signal is triggered.
[0012] Secondly, the present invention also provides a conveyor belt tear detection system based on a YOLACT network, comprising: Lasers, high-speed industrial cameras, and host computers; The laser is used to emit straight laser stripes onto the lower surface of the conveyor belt to be inspected; The high-speed industrial camera is used to acquire an image of the lower surface of the conveyor belt to be inspected; The host computer is used to extract the stripe center of the laser stripes in the lower surface image, obtain the stripe center line based on the stripe center, and input the stripe center line into the trained detection model to obtain the detection result output by the detection model; the detection model is obtained by training the YOLACT network based on the sample stripe center line and the sample label of the sample lower surface image; the sample label includes the tearing degree corresponding to the sample lower surface image.
[0013] Thirdly, the present invention also provides a conveyor belt tear detection device based on a YOLACT network, comprising: The acquisition unit is used to emit straight laser stripes onto the lower surface of the conveyor belt to be inspected using a laser, and acquire an image of the lower surface of the conveyor belt to be inspected. An extraction unit is used to extract the stripe center of the laser stripes in the lower surface image and obtain the stripe center line based on the stripe center; The detection unit is used to input the stripe center line into the trained detection model to obtain the detection result output by the detection model; the detection model is obtained by training the YOLACT network based on the sample stripe center line and the sample label of the sample lower surface image; the sample label includes the tearing degree corresponding to the sample lower surface image.
[0014] Fourthly, the present invention also provides an electronic device, including a memory and a processor, wherein, The memory is used to store programs; The processor, coupled to the memory, is used to execute the program stored in the memory to implement the conveyor belt tear detection method based on the YOLACT network as described in any of the above implementations.
[0015] Fifthly, the present invention also provides a computer-readable storage medium for storing a computer-readable program or instructions, which, when executed by a processor, can implement the steps in the conveyor belt tear detection method based on the YOLACT network described in any of the above implementations.
[0016] Sixthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps in the conveyor belt tear detection method based on the YOLACT network described in any of the above implementations.
[0017] The beneficial effects of this invention are as follows: The conveyor belt tear detection method and apparatus based on the YOLACT network provided by this invention emits straight laser stripes onto the lower surface of the conveyor belt to be inspected using a laser, and acquires an image of the lower surface of the conveyor belt. It utilizes the characteristic that the laser beam undergoes severe deformation at the cracks generated when the belt tears to determine whether the belt is torn. The center line of the laser stripes in the lower surface image is extracted to obtain the stripe centerline. Then, the YOLACT deep learning network is used to detect breakpoints in the stripe centerline. By leveraging the powerful feature learning and recognition capabilities of the YOLACT deep learning network, subtle changes in the laser stripes can be accurately detected. The detection accuracy for conveyor belt tears is far higher than that of traditional methods, and the detection results can be obtained accurately and quickly, improving the accuracy and reliability of conveyor belt tear detection. Attached Figure Description
[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0019] Figure 1 A schematic flowchart of an embodiment of the conveyor belt tear detection method based on YOLACT network provided by the present invention; Figure 2 A schematic diagram of an embodiment of the conveyor belt tear detection device based on the YOLACT network provided by the present invention; Figure 3 A schematic diagram of an embodiment of the electronic device provided by the present invention. Detailed Implementation
[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.
[0021] In the description of the embodiments of the present invention, unless otherwise stated, "multiple" means two or more. "And / or" describes the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.
[0022] The terms "first," "second," etc., used in the embodiments of this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a technical feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature.
[0023] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.
[0024] This invention provides a method and apparatus for detecting conveyor belt tearing based on the YOLACT network, which will be described below.
[0025] Figure 1 This is a schematic flowchart of an embodiment of the conveyor belt tear detection method based on the YOLACT network provided by the present invention, as shown below. Figure 1 As shown, the conveyor belt tear detection method based on the YOLACT network includes: S101. Use a laser to emit straight laser stripes onto the lower surface of the conveyor belt to be inspected, and acquire an image of the lower surface of the conveyor belt to be inspected. S102. Extract the center of the laser stripes in the lower surface image, and obtain the stripe center line based on the stripe center; S103. Input the stripe center line into the trained detection model to obtain the detection result output by the detection model; the detection model is obtained by training the YOLACT network based on the sample stripe center line of the sample lower surface image and the sample label of the sample lower surface image; the sample label includes the tearing degree corresponding to the sample lower surface image.
[0026] In step S101, a 650nm laser is used to project continuous straight laser stripes onto the bottom of the conveyor belt to be inspected, providing a comparable reference stripe for subsequent inspection. This laser has high stability and high brightness, enabling clear imaging under complex lighting conditions.
[0027] A high-speed industrial camera is used, whose frame rate and resolution meet the requirements for real-time acquisition of the laser beam status on the lower surface of the conveyor belt. The high-speed industrial camera can quickly capture image changes during the belt movement process and acquire images of the lower surface of the conveyor belt to be inspected in real time.
[0028] In S102, an image recognition algorithm is used to perform image recognition on the lower surface image, extract the stripe center of the laser stripe, and connect the stripe centers to obtain the stripe center line.
[0029] Specifically, a continuous straight laser stripe is projected onto the bottom of the belt by a laser. When the belt is running normally, the laser stripe on the lower surface of the belt is smooth, without jumps or breaks; however, when the belt tears, the red laser stripe will show breaks.
[0030] In S103, the detection model is obtained by training the YOLACT network in advance using the center line of the sample stripes in the sample underside image and the sample labels of the sample underside image. The sample labels include the degree of tearing corresponding to the sample underside image.
[0031] Leveraging the powerful feature learning and recognition capabilities of the YOLACT deep learning network, it can accurately detect subtle changes in laser stripes, achieving a much higher accuracy in detecting conveyor belt tears than traditional methods, effectively reducing false alarm and false negative rates.
[0032] By inputting the real-time acquired stripe center line into the detection model, the detection results output by the detection model are obtained. The detection results include whether the conveyor belt has been torn or the degree of tearing, such as the degree of tearing including minor tearing and severe tearing.
[0033] In summary, the conveyor belt tear detection method based on the YOLACT network provided in this invention uses a laser to emit straight laser stripes onto the lower surface of the conveyor belt to be inspected and acquires an image of the lower surface of the conveyor belt. It utilizes the characteristic that the laser beam causes severe deformation at the cracks created when the belt tears to determine whether the belt is torn. The center line of the laser stripes in the lower surface image is extracted to obtain the stripe centerline. Then, the YOLACT deep learning network is used to detect breakpoints in the stripe centerline. By leveraging the powerful feature learning and recognition capabilities of the YOLACT deep learning network, subtle changes in the laser stripes can be accurately detected. The detection accuracy for conveyor belt tears is far higher than that of traditional methods, and the detection results can be obtained accurately and quickly, improving the accuracy and reliability of conveyor belt tear detection.
[0034] In some embodiments of the present invention, the detection model is trained in the following manner: The degree of tearing on the lower surface image of the sample is marked to obtain a sample label; the sample label includes: no tear, slight tear, and severe tear; The YOLACT network is trained based on the center line of the sample stripes and the sample labels, with the minimum loss function as the learning objective, to obtain the detection model.
[0035] In some embodiments of the present invention, the expression for the loss function is as follows:
[0036] in, Represents the loss function. Indicates the actual value. Indicates the predicted value. m This indicates the number of images of the lower surface of the sample.
[0037] A large number of sample images of the conveyor belt's underside during normal operation and when it has torn are collected to construct a training dataset. The dataset should include samples of different types and degrees of tearing, as well as normal belt samples under various complex environments, to improve the model's generalization ability. The collected images are labeled to obtain sample labels, including no tear, slight tear, and severe tear, clearly marking the breakpoint of the laser stripes and information such as whether the belt has torn.
[0038] The YOLACT network model is trained using a labeled training dataset. During training, a loss function is constructed by combining the predicted values from the model output with the actual values from the training dataset, as follows:
[0039] in, Represents the loss function. Indicates the actual value. Indicates the predicted value. m This indicates the number of images of the lower surface of the sample.
[0040] By continuously adjusting the model's parameters and using the minimum loss function as the learning objective, the model can accurately identify abnormal changes in laser stripes and determine whether the conveyor belt has torn, thus obtaining a detection model.
[0041] Deep learning models are trained on a large number of samples in different environments and have good robustness to various interference factors.
[0042] The trained model is then validated and tested using a test dataset that was not used in the training process to evaluate its performance. By analyzing the test results, the model is further optimized, such as by adjusting its structure or adding more training data, until it reaches the expected detection accuracy and performance requirements.
[0043] The conveyor belt tear detection method based on the YOLACT network provided in this invention utilizes the powerful feature learning and recognition capabilities of the YOLACT deep learning network to accurately detect subtle changes in laser stripes. The detection accuracy for conveyor belt tears is far higher than that of traditional methods, effectively reducing false alarm and false negative rates.
[0044] In some embodiments of the present invention, after emitting straight laser stripes onto the lower surface of the conveyor belt to be inspected using a laser and acquiring an image of the lower surface of the conveyor belt to be inspected, the method further includes: The lower surface image is subjected to median filtering and CLAHE contrast enhancement processing.
[0045] Adaptive median filtering and CLAHE contrast enhancement processing are applied to the acquired lower surface image.
[0046] The adaptive median filter has a 5×5 window size and is used to eliminate salt-and-pepper noise. CLAHE contrast enhancement is used to increase the grayscale difference between the laser stripes and the background, making the grayscale level of the stripes higher than that of the background.
[0047] The conveyor belt tear detection method based on the YOLACT network provided in this invention improves image quality by performing preprocessing operations such as denoising and contrast enhancement on the acquired images, facilitating subsequent extraction of the laser stripe center. Filtering algorithms are used to remove noise interference from the image, and histogram equalization and other methods are employed to enhance the contrast between the laser stripes and the background.
[0048] In one possible implementation, after inputting the stripe center line into the trained detection model and obtaining the detection result output by the detection model, the method further includes: When the detection results indicate that the conveyor belt under test has torn, an alarm signal is triggered.
[0049] When the detection results indicate that the belt has torn, the alarm device is immediately triggered to notify the staff and send an emergency stop signal to the conveyor belt control system to prevent the accident from escalating further.
[0050] Alarm methods include audible and visual alarms, SMS notifications, and other forms to ensure that staff can receive alarm information in a timely manner.
[0051] The present invention also provides a conveyor belt tear detection system based on a YOLACT network, the system comprising: Lasers, high-speed industrial cameras, and host computers; The laser is used to emit straight laser stripes onto the lower surface of the conveyor belt to be inspected; The high-speed industrial camera is used to acquire an image of the lower surface of the conveyor belt to be inspected; The host computer is used to extract the stripe center of the laser stripes in the lower surface image, obtain the stripe center line based on the stripe center, and input the stripe center line into the trained detection model to obtain the detection result output by the detection model; the detection model is obtained by training the YOLACT network based on the sample stripe center line and the sample label of the sample lower surface image; the sample label includes the tearing degree corresponding to the sample lower surface image.
[0052] Using a 650nm laser beam as the detection light source, it has strong anti-interference capabilities and can work stably in complex industrial environments such as dust and light changes.
[0053] Install the laser at a suitable position at the bottom of the conveyor belt, ensuring that the laser stripes are projected evenly and completely onto the lower surface of the belt. Adjust the angle and position of the laser to achieve the best detection results.
[0054] A high-speed industrial camera is installed above the corresponding position of the laser to ensure that the camera's field of view can completely cover the laser stripe projection area.
[0055] Connect the host computer to the camera, laser, and alarm module to ensure stable data transmission lines between the host computer and each component, and avoid data loss or transmission interruption.
[0056] The host computer has built-in image processing software and a trained detection model. The image processing software preprocesses the image and extracts the center of the laser stripes in the lower surface image to obtain the stripe center line. The stripe center line is then input into the trained detection model to obtain the detection result output by the model.
[0057] The conveyor belt tear detection system based on the YOLACT network provided in this invention uses a laser to emit straight laser stripes onto the lower surface of the conveyor belt to be inspected, and acquires an image of the lower surface of the conveyor belt. It utilizes the characteristic that the laser beam causes severe deformation at the cracks created when the belt tears to determine whether the belt is torn. The center line of the laser stripes in the lower surface image is extracted to obtain the stripe centerline. Then, the YOLACT deep learning network is used to detect breakpoints in the stripe centerline. By leveraging the powerful feature learning and recognition capabilities of the YOLACT deep learning network, it can accurately detect subtle changes in the laser stripes. The detection accuracy for conveyor belt tears is far higher than that of traditional methods, and the detection results can be obtained accurately and quickly, improving the accuracy and reliability of conveyor belt tear detection.
[0058] In summary, this invention provides a conveyor belt tear detection method based on the YOLACT deep learning network. Utilizing the characteristic that severe deformation occurs at the cracks created when a laser beam tears a belt, the YOLACT deep learning network is used to determine whether a tear has occurred. A laser projects continuous straight laser stripes onto the bottom of the belt. After a high-speed camera acquires an image of the belt surface containing the laser stripes, a computer extracts the center of the stripes. The YOLACT deep learning network is then used to detect breakpoints in the center lines of the stripes. When the overall variation characteristics of the laser stripes exceed a set threshold, the belt is considered to have torn.
[0059] This invention utilizes a 650nm laser beam for high-speed, comprehensive scanning of the belt. A high-speed industrial camera captures the state of the laser beam on the lower surface of the belt. When the belt is operating normally without tearing, the laser stripes illuminating the lower surface of the belt are smooth, without jumps or breaks. When a tear occurs in the belt, the red laser stripes will show breaks. A visual tear detection algorithm is used to segment, extract, and analyze the features of the red laser stripes in the detection image to determine the belt condition, thereby achieving real-time monitoring, alarm, and emergency shutdown of the belt.
[0060] To better implement the conveyor belt tear detection method based on the YOLACT network in this invention embodiment, based on the YOLACT network-based conveyor belt tear detection method, correspondingly, as follows:Figure 2 As shown, this embodiment of the invention also provides a conveyor belt tear detection device based on a YOLACT network. The conveyor belt tear detection device 200 based on a YOLACT network includes: The acquisition unit 201 is used to emit straight laser stripes onto the lower surface of the conveyor belt to be inspected using a laser, and acquire an image of the lower surface of the conveyor belt to be inspected. Extraction unit 202 is used to extract the stripe center of the laser stripes in the lower surface image and obtain the stripe center line based on the stripe center; The detection unit 203 is used to input the stripe center line into the trained detection model to obtain the detection result output by the detection model; the detection model is obtained by training the YOLACT network based on the sample stripe center line and the sample label of the sample lower surface image; the sample label includes the tearing degree corresponding to the sample lower surface image.
[0061] The conveyor belt tear detection device 200 based on the YOLACT network provided in the above embodiments can realize the technical solutions described in the above embodiments of the conveyor belt tear detection method based on the YOLACT network. The specific implementation principles of each module or unit can be found in the corresponding content in the above embodiments of the conveyor belt tear detection method based on the YOLACT network, and will not be repeated here.
[0062] like Figure 3 As shown, the present invention also provides an electronic device 300. The electronic device 300 includes a processor 301, a memory 302, and a display 303. Figure 3 Only some components of the electronic device 300 are shown, but it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.
[0063] In some embodiments, processor 301 may be a central processing unit (CPU), microprocessor, or other data processing chip, used to run program code stored in memory 302 or process data, such as the conveyor belt tear detection method based on YOLACT network in this invention.
[0064] In some embodiments, processor 301 may be a single server or a group of servers. The server group may be centralized or distributed. In some embodiments, processor 301 may be local or remote. In some embodiments, processor 301 may be implemented on a cloud platform. In some embodiments, the cloud platform may include a private cloud, public cloud, hybrid cloud, community cloud, distributed cloud, internal cloud, multi-cloud, or any combination thereof.
[0065] In some embodiments, memory 302 may be an internal storage unit of electronic device 300, such as a hard disk or memory of electronic device 300. In other embodiments, memory 302 may also be an external storage device of electronic device 300, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. equipped on electronic device 300.
[0066] Furthermore, the memory 302 may include both internal storage units of the electronic device 300 and external storage devices. The memory 302 is used to store application software and various types of data installed on the electronic device 300.
[0067] In some embodiments, display 303 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an organic light-emitting diode (OLED) touchscreen. Display 303 is used to display information from electronic device 300 and to display a visual user interface. Components 301-303 of electronic device 300 communicate with each other via a system bus.
[0068] In one embodiment, when processor 301 executes a conveyor belt tear detection program based on a YOLACT network in memory 302, the following steps can be implemented: A laser is used to emit straight laser stripes onto the lower surface of the conveyor belt to be inspected, and an image of the lower surface of the conveyor belt to be inspected is acquired. Extract the center of the laser stripes in the lower surface image, and obtain the stripe center line based on the stripe center; The stripe centerline is input into the trained detection model to obtain the detection result output by the detection model; the detection model is obtained by training the YOLACT network based on the sample stripe centerline and the sample label of the sample lower surface image; the sample label includes the tearing degree corresponding to the sample lower surface image.
[0069] It should be understood that when the processor 301 executes the conveyor belt tear detection program based on the YOLACT network in the memory 302, in addition to the functions mentioned above, it can also perform other functions, as can be found in the description of the corresponding method embodiments above.
[0070] Furthermore, the embodiments of the present invention do not specifically limit the type of electronic device 300 mentioned. Electronic device 300 can be a mobile phone, tablet computer, personal digital assistant (PDA), wearable device, laptop computer, or other portable electronic device. Exemplary embodiments of portable electronic devices include, but are not limited to, portable electronic devices running iOS, Android, Microsoft, or other operating systems. The aforementioned portable electronic device can also be other portable electronic devices, such as a laptop computer with a touch-sensitive surface (e.g., a touch panel). It should also be understood that in some other embodiments of the present invention, electronic device 300 may not be a portable electronic device, but rather a desktop computer with a touch-sensitive surface (e.g., a touch panel).
[0071] Accordingly, embodiments of the present invention also provide a computer-readable storage medium for storing computer-readable programs or instructions. When the programs or instructions are executed by a processor, they can implement the steps or functions of the conveyor belt tear detection method based on the YOLACT network provided in the above-described method embodiments.
[0072] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to perform the steps or functions in the conveyor belt tear detection method based on the YOLACT network provided in the above-described method embodiments.
[0073] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware (such as a processor, controller, etc.), and the computer program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.
[0074] The conveyor belt tear detection method and device based on YOLACT network provided by the present invention have been described in detail above. Specific examples have been used to illustrate the principle and implementation of the present invention. The description of the above embodiments is only for the purpose of helping to understand the method and core idea of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation and application scope based on the idea of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.
Claims
1. A conveyor belt tear detection method based on YOLACT network, characterized in that, include: A laser is used to emit straight laser stripes onto the lower surface of the conveyor belt to be inspected, and an image of the lower surface of the conveyor belt to be inspected is acquired. Extract the center of the laser stripes in the lower surface image, and obtain the stripe center line based on the stripe center; The stripe centerline is input into the trained detection model to obtain the detection result output by the detection model; the detection model is obtained by training the YOLACT network based on the sample stripe centerline and the sample label of the sample lower surface image; the sample label includes the tearing degree corresponding to the sample lower surface image.
2. The conveyor belt tear detection method based on YOLACT network according to claim 1, characterized in that, The detection model was trained in the following way: The degree of tearing on the lower surface image of the sample is marked to obtain a sample label; the sample label includes: no tear, slight tear, and severe tear; The YOLACT network is trained based on the center line of the sample stripes and the sample labels, with the minimum loss function as the learning objective, to obtain the detection model.
3. The conveyor belt tear detection method based on YOLACT network according to claim 2, characterized in that, The expression for the loss function is as follows: in, Represents the loss function. Indicates the actual value. Indicates the predicted value. m This indicates the number of images of the lower surface of the sample.
4. The conveyor belt tear detection method based on YOLACT network according to claim 1, characterized in that, After emitting straight laser stripes onto the lower surface of the conveyor belt to be inspected using a laser and acquiring an image of the lower surface of the conveyor belt, the method further includes: The lower surface image is subjected to median filtering and CLAHE contrast enhancement processing.
5. The conveyor belt tear detection method based on YOLACT network according to claim 1, characterized in that, After inputting the stripe center line into the trained detection model and obtaining the detection result output by the detection model, the method further includes: When the detection results indicate that the conveyor belt under test has torn, an alarm signal is triggered.
6. A conveyor belt tear detection system based on a YOLACT network, characterized in that, include: Lasers, high-speed industrial cameras, and host computers; The laser is used to emit straight laser stripes onto the lower surface of the conveyor belt to be inspected; The high-speed industrial camera is used to acquire an image of the lower surface of the conveyor belt to be inspected; The host computer is used to extract the stripe center of the laser stripes in the lower surface image, obtain the stripe center line based on the stripe center, and input the stripe center line into the trained detection model to obtain the detection result output by the detection model; the detection model is obtained by training the YOLACT network based on the sample stripe center line and the sample label of the sample lower surface image; the sample label includes the tearing degree corresponding to the sample lower surface image.
7. A conveyor belt tear detection device based on a YOLACT network, characterized in that, include: The acquisition unit is used to emit straight laser stripes onto the lower surface of the conveyor belt to be inspected using a laser, and acquire an image of the lower surface of the conveyor belt to be inspected. An extraction unit is used to extract the stripe center of the laser stripes in the lower surface image and obtain the stripe center line based on the stripe center; The detection unit is used to input the stripe center line into the trained detection model to obtain the detection result output by the detection model; the detection model is obtained by training the YOLACT network based on the sample stripe center line and the sample label of the sample lower surface image; the sample label includes the tearing degree corresponding to the sample lower surface image.
8. An electronic device, characterized in that, Including memory and processor, among which, The memory is used to store programs; The processor, coupled to the memory, is used to execute the program stored in the memory to implement the steps in the conveyor belt tear detection method based on the YOLACT network as described in any one of claims 1 to 5.
9. A computer-readable storage medium, characterized in that, Used to store computer-readable programs or instructions, which, when executed by a processor, can implement the steps in the conveyor belt tear detection method based on the YOLACT network as described in any one of claims 1 to 5.
10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps in the conveyor belt tear detection method based on the YOLACT network as described in any one of claims 1 to 5.