Belt conveyor surface state detection method and device based on convolutional neural network

Abstract

The embodiment of the invention discloses a belt conveyor surface state detection method and device based on a convolutional neural network, and the method comprises the steps: collecting a large number of belt tearing, wearing and normal surface images, carrying out the image preprocessing of the collected belt surface images, adding an artificial label to the preprocessed belt surface image, randomly extracting more parts of belt surface images as a training set, inputting the training set into the established residual neural network to perform wear or tear feature extraction, and loading continuously input belt surface image features into a classifier to perform classification according to the extracted multi-dimensional feature value matrix so as to obtain a classification result; comparing a classification result with an artificial label, and returning an error rate and a loss function operation result to the residual neural network. and inputting the test set to obtain a final neural network model. By means of the model, the abrasion or tearing recognition operation can be executed on the belt surface image shot in real time, testing is conducted through the testing set, false detection is removed, and the abnormal state recognition accuracy is improved.

Description

technical field [0001] The embodiment of the present application relates to the technical field of material transportation equipment detection, and in particular to a method and device for detecting the surface state of a belt conveyor based on a convolutional neural network. Background technique [0002] The belt conveyor is a machine that continuously transports materials through friction drive. The materials can move from one end of the feeding point to the other end of the belt along the direction of belt conveying, and finally reach the unloading point to complete the material conveying process. Among the main components of the belt conveyor, the belt is not only the bearing part of the material, but also provides the friction force for the material to move, so the surface quality of the belt directly affects the material transportation effect. [0003] During the process of conveying materials, materials, impurities or metal parts are easy to cause wear on the belt sur...

AI Technical Summary

Problems solved by technology

[0005] This application provides a method for detecting the surface state of belt conveyors based on convolutional neural networks to solve the problems that existing methods cannot implement real-time detection of belt surface states and cannot predict belt breakage. The advanced detection method of the surface condition of the belt conveyor can timely and effectively issue an early warning when the belt is worn, so as to timely remedy the belt, prolong the service life of the belt, and reduce the economic loss of the factory

Images