CN116433645A - Belt bulge detection method and detection system - Google Patents
Belt bulge detection method and detection system Download PDFInfo
- Publication number
- CN116433645A CN116433645A CN202310438944.6A CN202310438944A CN116433645A CN 116433645 A CN116433645 A CN 116433645A CN 202310438944 A CN202310438944 A CN 202310438944A CN 116433645 A CN116433645 A CN 116433645A
- Authority
- CN
- China
- Prior art keywords
- bulge
- module
- belt
- image
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 62
- 238000012549 training Methods 0.000 claims abstract description 39
- 230000005540 biological transmission Effects 0.000 claims abstract description 24
- 238000012545 processing Methods 0.000 claims abstract description 12
- 238000003860 storage Methods 0.000 claims abstract description 10
- 230000008859 change Effects 0.000 claims abstract description 6
- 230000002159 abnormal effect Effects 0.000 claims description 41
- 238000000034 method Methods 0.000 claims description 17
- 238000005520 cutting process Methods 0.000 claims description 3
- 125000004122 cyclic group Chemical group 0.000 claims description 3
- 230000002708 enhancing effect Effects 0.000 claims description 3
- 230000006870 function Effects 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims 1
- 238000004519 manufacturing process Methods 0.000 abstract description 4
- 238000004891 communication Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 238000007726 management method Methods 0.000 description 4
- 238000010295 mobile communication Methods 0.000 description 4
- 230000007246 mechanism Effects 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 238000005272 metallurgy Methods 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 229910000831 Steel Inorganic materials 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000004566 building material Substances 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 239000003208 petroleum Substances 0.000 description 1
- -1 power plants Substances 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 239000010959 steel Substances 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
- G06V10/765—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects using rules for classification or partitioning the feature space
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30164—Workpiece; Machine component
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Abstract
A belt bulge detection method and a detection system, in particular to a belt bulge detection method and a belt bulge detection system based on a 5G and CV pre-training large model. The detection method comprises the following steps: opening a dustproof camera module to acquire field image data of the belt surface; transmitting the image to a video service module through a 5G transmission module for video processing and storage; the model identification module pulls the real-time video stream to carry out detection and identification; the system circularly and intelligently judges whether the detection result is bulge or not based on the extracted image; the alarm module autonomously judges whether the bulge is the original bulge without change, if the detection result is a new bulge, the alarm is carried out, otherwise, the alarm is not carried out. The invention can accurately identify the belt bulge and continuously monitor the change of the belt bulge, thereby reducing the influence of the belt bulge on the production process.
Description
Technical Field
The invention belongs to the technical field of belt conveyors, relates to a belt bulge detection method and a detection system, and particularly relates to a belt bulge detection method and a belt bulge detection system based on a 5G and CV pre-training large model.
Background
The belt conveyor is an important conveying device for conveying materials in enterprises in various fields such as ports, metallurgy, mines, chemical industry, petroleum, power plants, building materials and the like. Due to advances in technology, belt conveyors continue to increase in speed, scale, and transport distance. The existing steel enterprises have a large number of articles in the bulk cargo transportation process of production, and the belt conveyor needs long-time and high-load transportation, so that the related range is very wide. Therefore, the bulge of the conveyor belt becomes an important index for measuring the safety of the conveyor belt, a detection system capable of continuously detecting the bulge state and change of the conveyor belt is researched, an alarm is accurately and timely sent out, the influence of the bulge of the conveyor belt on the production process is reduced to the greatest extent, and great economic and social benefits are achieved.
The fifth generation mobile communication technology (5 th Generation Mobile Communication Technology, abbreviated as 5G) is a new generation broadband mobile communication, has the characteristics of high speed, low time delay and large connection, not only solves the problem of human-to-human communication, but also solves the problem of human-to-object communication, and meets the communication requirements in the fields of industrial control, intelligent manufacturing and the like. The peak rate of 5G can reach more than 11Gbit/s at present, so that the high-data-volume transmission of high-definition videos and the like is satisfied, the air interface time delay is as low as 1ms, and the requirements of real-time monitoring and control are satisfied; meanwhile, the device has the device connection capability of million connections/square kilometer, and meets the communication requirement of the Internet of things. Aiming at complex environments such as ports, metallurgy, mines, chemical industry and the like, the deployment difficulty is greatly reduced and the time is shortened by using 5G communication compared with the traditional wiring mode.
The AI large Model is a Foundation Model, which is a Model that can adapt to a series of downstream tasks by training on large-scale broad data. The appearance of the pre-training large model is a necessary result of combining artificial intelligence with big data and big calculation force, and is a milestone technology of the artificial intelligence for general intelligence. The large AI model is oriented to actual tasks, training is performed on massive general data before modeling, knowledge is automatically learned from the general data based on the supervised neural network and the self-supervised neural network, and then the large AI model is performed on a specific scene based on a small amount of special data, so that the contradiction between the rapid increase of the general data and the lack of the special data of the AI model in the industrial field is solved, and the generalization, the universality and the practicability of the AI are greatly improved. The CV pre-training large model is a visual AI large model trained based on massive image data.
Disclosure of Invention
The invention aims to provide a belt bulge detection method and a detection system based on a 5G and CV pre-training large model, which improve the recognition accuracy of bulge, thereby accurately recognizing the bulge of the belt and timely alarming and prompting the operation of staff.
The invention realizes the above purpose through the following technical scheme:
a belt bulge detection method comprises the following steps:
s1, opening a dustproof camera module to acquire field image data of a belt surface;
s2, transmitting the image to a video service module through a 5G transmission module for video processing and storage;
s3, the model identification module pulls the real-time video stream, frames are taken according to the set frequency, and the frames are sent to the belt bulge detection model for detection and identification;
s4, the system circularly and intelligently judges whether the detection result is a bulge or not based on the extracted image, and if the detection result is the bulge, a bulge abnormal message is sent, wherein the bulge abnormal message comprises an abnormal picture, bulge coordinates, size and the like to an alarm module; if the detection result is not bulge, a bulge abnormal message is not sent.
S5, the warning module autonomously judges whether the bulge is the original bulge without change or not: if the detection result is a new bulge, alarming is carried out to remind workers to operate according to the specification; if the detection result is the original bulge, no alarm is given.
The further scheme is as follows: in the step S2, the 5G transmission module includes a 5G CPE, and the 5G CPE converts the video data collected by the front end camera module into a 5G wireless signal for rapid transmission;
the further scheme is as follows: in the step S3, the system further includes a picture processing module, where the picture processing module includes: selecting an image area, namely cutting an original image according to a set area, selecting only an image of a belt surface part, reducing the calculated amount, and avoiding false identification and alarm caused by surrounding environment images; and (3) enhancing the image, namely improving the brightness and contrast of the image through histogram equalization and improving the image quality.
The further scheme is as follows: in the step S3, the bulge identifying model step includes: creating a CV large model pre-training data set, creating a model learning data set, and training and building a model by using the data set; the model building method comprises the following steps of:
firstly, collecting a large number of data sets of the normal state of the belt and a small number of the bulge state of the belt from the site, carrying out local storage, and uploading the data sets to a cloud model training module;
the method is characterized by comprising the steps of adopting an image classification method to extract characteristics of normal and abnormal images, wherein the characteristics of the normal and abnormal images comprise information such as size, shape, gray value, RGB value and the like, and distinguishing the normal image from the abnormal image;
the method is characterized by comprising the steps of extracting the characteristics of an abnormal image by adopting a Faster R-CNN target detection method, wherein the characteristics comprise information such as size, shape, gray value, RGB value and the like, adding a target frame for a target area, and giving out recognition confidence.
The further scheme is as follows: in the step S3, the detection and identification result of the belt bulge model includes: the system identifies and decides the bulge state of the belt conveyor;
and for the picture data transmitted from the camera module, carrying out decision on the system according to an accuracy rate result after the picture data is identified by the belt bulge identification model: if the accuracy rate is above 90%, sending a bulge abnormal message to the alarm module, wherein the bulge abnormal message comprises an abnormal picture, bulge coordinates and size; if the detection result is not bulge, no abnormal message is sent.
The further scheme is as follows: in the step S5, the alarm module includes whether the original bulge is unchanged, and determining: judging whether the bulge is at the same position or not based on the bulge coordinates and the belt speed by adopting a cyclic comparison method, comparing the bulge sizes, judging whether the same bulge is changed or not, and if both conditions are met, carrying out alarm prompt; otherwise, the alarm prompt is not carried out.
The further scheme is as follows: in step S5, the step of storing the image determined to be new bulge as a data set image for learning and using at the same time includes storing image data of the belt bulge state in the local area by adopting a method for learning and using at the same time, collecting the manually confirmed non-bulge abnormal image, uploading the image to a cloud training module for model learning and updating, and issuing the updated model to a local model identification module on line.
A belt bulge detection system, comprising:
the data acquisition module is used for acquiring field image data of the belt conveyor by adopting a dustproof camera shooting module;
the 5G transmission module is used for carrying out large-bandwidth wireless rapid image transmission to the video service module by adopting 5G for image processing and storage;
the video service module provides functions of accessing, managing and storing data of the front-end camera module, and supports unified grouping management of the camera module, including accessing, restarting and the like. The working efficiency can be greatly improved through grouping management of the camera modules; the fault-tolerant mechanism is provided based on the continuity, the instantaneity and the time sequence of the streaming media, so that the high reliability of data storage is ensured;
the cloud model training module is used for creating a data set of normal and bulge states of the belt and training a belt bulge identification model through a CV pre-training large model;
the local model recognition module is used for recognizing the bulge state of the real-time belt image by using the belt bulge recognition model obtained in the model training module and making a relevant decision on the obtained result;
and the alarm decision module is used for deciding whether the bulge is unchanged or not by comparing bulge data in the newly transmitted abnormal message with original bulge data, if the bulge is unchanged, the alarm decision module does not transmit an alarm prompt, and otherwise, the alarm decision module sends an alarm through an alarm.
The data acquisition module is connected to the 5G transmission module through a wired network; the 5G transmission module transmits the data to the 5G base station through a wireless network; the 5G base station is connected with the video service module through a wired network; the video service module, the local model identification module, the cloud model training module and the alarm decision module are connected through a network.
The beneficial effects of the invention are as follows: the belt bulge identification model obtained through training can automatically distinguish whether the belt bulge is normal, learn the characteristics of the belt bulge, update the iteration model in real time through a decision mechanism of the system, improve the identification accuracy of the bulge, accurately identify the bulge of the belt, and timely alarm and prompt the operation of staff.
Drawings
FIG. 1 is a schematic flow chart of the present invention.
FIG. 2 is a schematic diagram of a system module according to the present invention.
Fig. 3 is a schematic view of the field structure of the present invention.
In the figure: 1. the system comprises a belt conveyor, a dustproof camera module, a video service module, a model identification module, an alarm module and a cloud training module.
Detailed Description
The invention is further described below with reference to the accompanying drawings:
referring to fig. 1 to 3, a belt bulge detection method based on a 5G and CV pre-training large model includes the steps of:
s1, opening a dustproof camera module (2) to acquire field image data of a belt surface (1);
s2, transmitting the image to a video service module (3) through a 5G transmission module for video processing and storage;
s3, a model identification module (4) pulls the real-time video stream, frames are taken according to the set frequency, and the frames are sent to a belt bulge detection model for detection and identification;
s4, the system intelligently judges whether the detection result is a bulge or not based on the extracted image in a circulating way; if the detection result is a bulge, a bulge abnormal message is sent to an alarm module (5) including an abnormal picture, bulge coordinates, size and the like; if the detection result is not bulge, not sending bulge abnormal information;
s5, the warning module autonomously judges whether the bulge is the original bulge without change or not: if the detection result is a new bulge, alarming is carried out to remind workers to operate according to the specification; if the detection result is the original bulge, no alarm is given.
In the embodiment of the present invention, in the step S2, the 5G transmission module includes: the 5G CPE and 5G are fifth generation mobile communication technologies, have the characteristics of high speed, low time delay and large connection, allow users to carry out rapid image transmission based on a wireless connection mode, convert video data collected by a front-end camera module into 5G wireless signals for rapid transmission, and greatly reduce deployment difficulty and time under complex industrial environments without laying optical fibers.
In the embodiment of the present invention, in the step S3, a picture processing module is further included, and the steps include: selecting an image area, namely cutting an original image according to a set area, selecting only an image of a belt surface part, reducing the calculated amount, and avoiding false identification and alarm caused by surrounding environment images; and (3) enhancing the image, namely improving the brightness and contrast of the image through histogram equalization and improving the image quality.
In the embodiment of the present invention, in the step S3, the bulge identifying model step includes: creating a CV large model pre-training data set, creating a model learning data set, and training and building a model by using the data set; the method comprises the following steps of: firstly, collecting a large number of data sets of the normal state of the belt and a small number of the bulge state of the belt from the site, carrying out local storage, and uploading the data sets to a cloud model training module; the method is characterized by comprising the steps of adopting an image classification method to extract characteristics of normal and abnormal images, wherein the characteristics of the normal and abnormal images comprise information such as size, shape, gray value, RGB value and the like, and distinguishing the normal image from the abnormal image; the method is characterized by comprising the steps of extracting the characteristics of an abnormal image by adopting a Faster R-CNN target detection scheme, wherein the characteristics comprise information such as size, shape, gray value, RGB value and the like, adding a target frame for a target area, and giving out recognition confidence.
In the embodiment of the present invention, in the step S3, the detection and identification result of the belt bulge model includes: the system identifies and decides the bulge state of the belt conveyor; and for the picture data transmitted from the camera module, carrying out decision on the system according to an accuracy rate result after the picture data is identified by the belt bulge identification model: if the accuracy rate is above 90%, sending a bulge abnormal message to the alarm module, wherein the bulge abnormal message comprises an abnormal picture, bulge coordinates, size and the like; if the detection result is not bulge, no abnormal message is sent.
In the embodiment of the present invention, in the step S5, the alarm module includes an identification and decision of whether the original bulge is unchanged: judging whether the bulge is at the same position or not based on the bulge coordinates and the belt speed by adopting a cyclic comparison method, comparing the bulge sizes, judging whether the same bulge is changed or not, and if both conditions are met, carrying out alarm prompt; otherwise, the alarm prompt is not carried out.
In the embodiment of the present invention, in the step S5, the step of storing the picture determined to be the new bulge as the picture of the data set for learning while includes: the method for learning and using simultaneously is adopted, image data of a belt bulge state is stored locally, manually confirmed non-bulge abnormal pictures are collected and uploaded to a cloud training module for model learning and updating, and the updated model is issued to a local model identification module on line.
Referring to fig. 1 to 3, a belt bulge detection system based on a 5G and CV pre-training large model, comprising:
the data acquisition module is used for acquiring field image data of the belt conveyor (1) by adopting the dustproof camera module (2);
the 5G transmission module is used for carrying out large-bandwidth wireless rapid image transmission to the video service module (3) by adopting 5G for image processing and storage;
the video service module provides functions of accessing, managing and storing data of the front-end camera module, and supports unified grouping management of the camera module, including accessing, restarting and the like. The working efficiency can be greatly improved through grouping management of the camera modules; and a fault-tolerant mechanism is provided based on the continuity, the instantaneity and the time sequence of the streaming media, so that the high reliability of data storage is ensured.
And the cloud model training module is used for creating a data set of normal and bulge states of the belt and training a belt bulge identification model through the CV pre-training large model.
The local model recognition module is used for recognizing the bulge state of the real-time belt image by using the belt bulge recognition model obtained by the model training module, and carrying out relevant decision on the obtained result: if the decision result is a bulge, an abnormal message is sent, and if the decision result is no, no abnormal message is sent.
The alarm decision module is used for deciding whether the original bulge is unchanged or not by comparing bulge data in the newly transmitted abnormal message with original bulge data; if the original bulge is unchanged, an alarm prompt is not sent; otherwise, an alarm is sent out through an alarm.
The data acquisition module is connected to the 5G transmission module through a wired network; the 5G transmission module transmits the data to the 5G base station through a wireless network; the 5G base station is connected with the video service module through a wired network; the video service module, the local model identification module, the cloud model training module and the alarm decision module are connected through a network.
Working principle: the on-site image is collected by adopting a dustproof camera, images of normal and bulge of the belt are trained by using a CV pre-training large model, and bulge identification is carried out on the target image by outputting real-time images.
Claims (8)
1. The belt bulge detection method is characterized by comprising the following steps of:
s1, opening a dustproof camera module to acquire field image data of a belt surface;
s2, transmitting the image to a video service module through a 5G transmission module for video processing and storage;
s3, the model identification module pulls the real-time video stream, frames are taken according to the set frequency, and the frames are sent to the belt bulge detection model for detection and identification;
s4, the system circularly and intelligently judges whether the detection result is a bulge or not based on the extracted image, and if the detection result is the bulge, the bulge abnormal message is sent to the alarm module, wherein the bulge abnormal message comprises an abnormal picture, bulge coordinates and size information; if the detection result is not bulge, not sending bulge abnormal information;
s5, the warning module autonomously judges whether the bulge is the original bulge without change or not: if the detection result is a new bulge, alarming is carried out to remind workers to operate according to the specification; if the detection result is the original bulge, no alarm is given.
2. The belt bulge detection method as set forth in claim 1, wherein: in step S2, the 5G transmission module includes a 5G CPE, and the 5G CPE converts the video data collected by the front end camera module into a 5G wireless signal for quick transmission.
3. The belt bulge detection method as set forth in claim 1, wherein: in the step S3, a picture processing module is further included; the picture processing module comprises: selecting an image area, namely cutting an original image according to a set area, selecting only an image of a belt surface part, reducing the calculated amount, and avoiding false identification and alarm caused by surrounding environment images; and (3) enhancing the image, namely improving the brightness and contrast of the image through histogram equalization and improving the image quality.
4. The belt bulge detection method as set forth in claim 1, wherein: in the step S3, the bulge identifying model step includes: creating a CV large model pre-training data set, creating a model learning data set, and training and building a model by using the data set; the model building method comprises the following steps of:
firstly, collecting a large number of data sets of the normal state of the belt and a small number of the bulge state of the belt from the site, carrying out local storage, and uploading the data sets to a cloud model training module;
adopting an image classification method to extract characteristics of normal and abnormal images including size, shape, gray value and RGB value information, and distinguishing the normal image from the abnormal image;
and (3) performing feature extraction on the abnormal image including size, shape, gray value and RGB value information by adopting a Faster R-CNN target detection method, adding a target frame for a target area, and giving out recognition confidence.
5. The belt bulge detection method as set forth in claim 1, wherein: in the step S3, the detection and identification result of the belt bulge model includes identification and decision of the system on the bulge state of the belt conveyor: after the image data transmitted from the camera module is identified by the belt bulge identification model, judging that the accuracy is more than 90%, and transmitting bulge abnormal information, including abnormal images, bulge coordinates and sizes, to the alarm module; if the detection result is not bulge, no abnormal message is sent.
6. The belt bulge detection method as set forth in claim 1, wherein: in the step S5, the alarm module includes whether the original bulge is unchanged, and determining: judging whether the bulge is at the same position or not based on the bulge coordinates and the belt speed by adopting a cyclic comparison method, comparing the bulge sizes, judging whether the same bulge is changed or not, and if both conditions are met, carrying out alarm prompt; otherwise, the alarm prompt is not carried out.
7. The belt bulge detection method as set forth in claim 1, wherein: in step S5, the step of storing the image determined to be new bulge as a data set image for learning and using at the same time includes storing image data of the belt bulge state in the local area by adopting a method for learning and using at the same time, collecting and uploading manually confirmed non-bulge abnormal images to a cloud training module for model learning and updating, and issuing the updated model to a local model identification module on line.
8. A belt bulge detection system, characterized in that the detection system comprises:
the data acquisition module is used for acquiring field image data of the belt conveyor by adopting a dustproof camera shooting module;
the 5G transmission module is used for carrying out large-bandwidth wireless rapid image transmission to the video service module by adopting 5G for image processing and storage;
the video service module provides functions of accessing, managing and storing data of the front-end camera module, and supports unified grouping management of the camera module, including accessing and restarting;
the cloud model training module is used for creating a data set of normal and bulge states of the belt and training a belt bulge identification model through a CV pre-training large model;
the local model recognition module is used for recognizing the bulge state of the real-time belt image by using the belt bulge recognition model obtained in the model training module and making a relevant decision on the obtained result;
the alarm decision module is used for deciding whether the original bulge is unchanged or not by comparing bulge data in the newly transmitted abnormal message with original bulge data;
the data acquisition module is connected to the 5G transmission module through a wired network; the 5G transmission module transmits the data to the 5G base station through a wireless network; the 5G base station is connected with the video service module through a wired network; the video service module, the local model identification module, the cloud model training module and the alarm decision module are connected through a network.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310438944.6A CN116433645A (en) | 2023-04-23 | 2023-04-23 | Belt bulge detection method and detection system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310438944.6A CN116433645A (en) | 2023-04-23 | 2023-04-23 | Belt bulge detection method and detection system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116433645A true CN116433645A (en) | 2023-07-14 |
Family
ID=87085229
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310438944.6A Pending CN116433645A (en) | 2023-04-23 | 2023-04-23 | Belt bulge detection method and detection system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116433645A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116758400A (en) * | 2023-08-15 | 2023-09-15 | 安徽容知日新科技股份有限公司 | Method and device for detecting abnormality of conveyor belt and computer readable storage medium |
-
2023
- 2023-04-23 CN CN202310438944.6A patent/CN116433645A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116758400A (en) * | 2023-08-15 | 2023-09-15 | 安徽容知日新科技股份有限公司 | Method and device for detecting abnormality of conveyor belt and computer readable storage medium |
CN116758400B (en) * | 2023-08-15 | 2023-10-17 | 安徽容知日新科技股份有限公司 | Method and device for detecting abnormality of conveyor belt and computer readable storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111582016A (en) | Intelligent maintenance-free power grid monitoring method and system based on cloud edge collaborative deep learning | |
CN210691345U (en) | Cloud intelligence safety inspection auxiliary system based on degree of depth learning | |
CN110996067A (en) | Personnel safety real-time intelligent video monitoring system under high-risk operation environment based on deep learning | |
CN110516529A (en) | It is a kind of that detection method and system are fed based on deep learning image procossing | |
CN102143354B (en) | Method for recognizing and calculating galloping of transmission conductor based on video image processing | |
CN112149522A (en) | Intelligent visual external-damage-prevention monitoring system and method for cable channel | |
CN116433645A (en) | Belt bulge detection method and detection system | |
CN104079885A (en) | Nobody-monitored and linkage-tracked network camera shooting method and device | |
CN111223263A (en) | Full-automatic comprehensive fire early warning response system | |
CN111598855B (en) | 2C equipment high-speed rail contact net dropper defect detection method based on deep learning and transfer learning | |
CN113642403A (en) | Crowd abnormal intelligent safety detection system based on edge calculation | |
CN115457446A (en) | Abnormal behavior supervision system based on video analysis | |
CN113516102A (en) | Deep learning parabolic behavior detection method based on video | |
CN116846059A (en) | Edge detection system for power grid inspection and monitoring | |
CN219340740U (en) | Belt bulge detecting system | |
CN112686130B (en) | Wisdom fishing boat supervision decision-making system | |
Daogang et al. | Anomaly identification of critical power plant facilities based on YOLOX-CBAM | |
CN117423157A (en) | Mine abnormal video action understanding method combining migration learning and regional invasion | |
CN112232235A (en) | Intelligent factory remote monitoring method and system based on 5G | |
Xie et al. | A package auto-counting model based on tailored YOLO and DeepSort techniques | |
CN114592411B (en) | Carrier parasitic type intelligent inspection method for highway damage | |
CN116248861A (en) | Intelligent video detection method, system and device | |
CN115393900A (en) | Intelligent construction site safety supervision method and system based on Internet of things | |
CN103345625A (en) | Method and system for analyzing three-dimensional image | |
CN207399404U (en) | A kind of tracking monitoring system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |