Computer Vision-based Detection Technology And Application Of Load Degree Of Open-pit Mine Trucks

Truck load statistical are related to open-pit mine’s production schedule,operation arrangements,production dispatching efficiency,etc.And truck load statistical directly affect the economic benefits of the mine,which is of great significance to mine production management.How to improve the accuracy and efficiency of truck load statistics always is a problem that need to be solved but still unresolved in mine production management.For reducing the cost while improving work efficiency and arousing the initiative of production staff,the automatic detection of truck load degree is necessary to realize optimization of truck load statistical management methods,to speed up the progress of mine information construction,and to realize enterprise benefit maximization.The main contents of the study are as follows:(1)Guided by the prerequisites that need to be met for image feature extraction of open-pit mine truck load degree,starting from the problems existing in truck load degree image data sets,a method for image pre-processing of truck load degree is proposed,including data sets balancing method,data augmentation method and image downsampling method.(2)Based on the results of truck load degree image pre-processing,an image feature extraction method for open pit mine truck load degree image is proposed,including truck load degree image feature extraction methods based on LBP and Alex Net.The rotation invariance uniform pattern LBP and second order LDP feature maps of the truck load degree image are fused with the RGB pixel value feature maps,and then the deep features of the truck load degree image are extracted from the fused features based on the Alex Net.(3)Based on the results of the truck load degree image feature extraction,an open-pit mine truck load degree detection method is proposed,and an open-pit mine truck load degree detection model is constructed.The parameters of the open-mine truck load degree detection model are selected and the truck load degree detection model is solved.(4)Based on the research on the load degree detection technology of open-pit mine truck,construct a computer vision-based truck load degree detection system,complete the design of the system architecture,functional modules and deployment scenarios,and analyzes the effectiveness of the proposed truck load degree detection technology and system based on the implementation and application of the truck load degree detection technology and system in this mine.The results of system application analysis in the M open-pit mine show that the application of the truck load degree detection technology and the truck load degree detection system can realize the upgrade of the truck load statistical management method,which has certain application value.