Design Of Target Tracking System Based On Embedded

In recent years,with the development of image processing,pattern recognition,target tracking,and the development of low power consumption and high performance processors.Embedded visual products are widely used in intelligent transportation,intelligent video surveillance,robot vision navigation,virtual reality,unmanned,unmanned aerial vehicles and other fields,which constantly improve and affect people's lives,so that people's lives become more convenient and intelligent.The introduction of embedded machine vision technology provides technical accumulation for mobile devices and machines which need real-time processing,so that the machine's "eye" real-time better,faster response.In this paper,a set of target tracking system based on embedded system is designed by combining the target tracking algorithm with embedded system and dedicated HD video processor.The experimental test satisfies the functional requirements.The main work done in this paper is as follows:(1)According to the functional requirements of the system,the hardware framework is designed,and the hardware platform is built.Front end selects the Panasonic GP_MH322 model to collect 1080*1920 HD video.Then,DaVinci TMS320DM8127 HD video processing SOC is used to process the collected video as well as the control scheduling of the whole system.Nand Flash as a system to run,start the code storage,DDR3 storage video and other massive data.Finally,through the computer browser to access the WEB server display video.In addition,in order to expand the system monitoring field of vision,selected Ming King TK-35 model PTZ.(2)Development of system software development environment.Install embedded software development commonly used tools and configure the virtual machine linux system,including tftp server configuration,NFS server configuration.Install and analyze the IPNC RDK DaVinci software development kit,and transplant the embedded linux system into the development board according to the hardware platform,including uboot transplantation,kernel transplantation,file system transplantation.(3)The goal tracking and PTZ control applications were developed by analyzing the multi-channel software architecture and the DaVinci multi-core communication principle.Based on the benchmark criterion,the target tracking algorithm is compared and analyzed by Matlab simulation.Finally,STC is used as the tracking algorithm of the system,and the tracking algorithm is rewritten into the system by C/C++ language.Based on the PTZ communication protocol,the PTZ control method is designed according to the relationship between the center position of the image tracking frame and the rotational speed of the pan.Finally,the system test,including the system start test,tracking algorithm test,PTZ control test,verify the feasibility of the system.