Video Image Processing Based Belt Deviation Detection Algorithm And Its Application

Belt conveyor is a continuous transportation machine widely used in coal production,housing construction,material metallurgy,power supply and other industries^[1].During the operation of the belt conveyor,the most common fault of the belt conveyor is the deviation of the belt.If it is not handled in time,it is easy to increase the number of conveyor accidents and affect normal factory production operations,and it is very easy to cause materials to spill out,which reduces the operating economy of the transportation system,shortens the service life of the machine,and affects production efficiency.Have undesirable consequences such as security risks^[2].Traditional belt deviation detection requires manual real-time monitoring of the video,which consumes a lot of manpower and has low accuracy;even if the belt deviation is manually detected,it may not be able to early warning and processing in time.The belt deviation detection based on video image processing can save a lot of manpower and material resources,find the deviation problem in time,effectively and accurately,and effectively warn the deviation of the belt at the first time and deal with it,so as to minimize the defects caused by hidden safety hazards.as a result of.This article first studies and analyzes the belt deviation problem.A more detailed analysis of the cause of belt deviation and deviation detection is carried out;on this basis,a belt deviation algorithm based on video images is proposed.The algorithm undergoes video drainage,frame extraction,grayscale transformation,edge detection,and pixel After point fitting and other processing,it can be compared with the pre-marked belt benchmark to judge the deviation,and can be marked and early warning on the original video image.In the edge detection algorithm,after experimenting,analyzing and comparing the existing Roberts operator,Prewitt operator,Sobel operator,Scharr operator,Laplacian operator and Log operator,the Canny with the best comprehensive performance is finally selected.The operator is used as the belt edge detection algorithm;then combined with the actual belt video samples of Jinyu Jidong Cement Enterprise,the proposed belt deviation detection algorithm is tested and analyzed.The results show that the belt deviation detection algorithm proposed in this paper is compared with simple The Canny operator,template edge detection and other image processing methods have better accuracy and robustness.Because the edge extraction algorithm is affected by the environment,especially in the case of serious outdoor light pollution,the effect is unstable;the algorithm of this research has carried out gray-scale conversion,which can better avoid the influence of factors such as light and weather;at the same time,relative Compared with algorithms such as support vector machine（SVM）and deep learning,the detection algorithm proposed in this paper does not require a large number of training samples and training time,saving a large amount of sample collection costs and training time costs,and has the advantages of accuracy and time complexity.Compared with the current belt deviation detection,it has obvious advantages.Based on the proposed algorithm,the author combined the actual needs of the engineering project,followed the process,principles and methods of software engineering to complete the design and implementation of the belt running detection and early warning system of BBMG Jidong Cement Guangling Company.The proposed algorithm is applied to belt deviation detection,and real-time warning can be implemented for the detected deviation,so as to correct the belt deviation problem in time and reduce the subsequent safety hazards caused by belt deviation.