| Information monitoring of slag landfill site is of great importance to environmental protection.Iron ore slag will occupy a large amount of land and poisoning the soil if dump carelessly becouse the basicity production residue of steelworks and has residual flotation reagent ion and heavy metals.Usually adopted construction of iron ore slag landfill to store slag and use an intelligent monitoring system to ensure safe and environmental operation of slag landfill site.It is of practical significance to design an intelligent monitoring system for iron ore slag landfill site.This dissertation is the requirement of engineering application exploited an iron ore slag landfilling intelligent monitoring system.system developed with B/S and C/S mixed architecture could achieve the query of the dumping behavior of slag trucks entry and exit with the coordinate of real-time video collection system.The main research works are as follows:A slag landfill real-time video collection system was set up for three parts,inbound and outbound acquisition cameras,mine acquisition camera arrays and communication networks.Inbound and outbound cameras was set up to obtain time information and vehicle information.Slag landfill site real-time video streams was set up with an array of mine cameras and connection with optical communication network.A C/S Architecture Service layer loftware system has been built with c# multithreading technology to achieve the deep learning recognition and data processing.A B/S architecture terminal data rendering system was developed by ASP.NET MVC frame,through a fixed address to accesss.Users sign in the system with account and passwords.An improved YOLOv5 identifying dumping status of slag trucks algorithm was proposed with complete deep learning and training,the recognition accuracy of vehicle behavior reaches 90.1%.The practical application was shown that system runs stably and can effectively complete digitized monitoring and intelligent identification of the slag truck working process,it has engineering application value for improving the information and intelligence in many different scenarios. |