Design And Research Of Moving Target Tracking System Based On FPGA

With the rapid development of electronic technology, computer performance has been greatly improved. Computer realization of human visual function is becoming one of the most popular topics in the computer field. Target detection and tracking technology based on video is the most important field of computer vision research. It is the key technology in the industrial applications such as intelligent control, human-computer interaction, mobile robot navigation and robot hand-eye system. It has very attractive prospect in scientific research and engineering applications.This dissertation proposes a new idea based on FPGA, designs a static background video sequence used in the dynamic object detection and tracking system. A certain region is monitored by a fixed camera, the dynamic video sequences are analyzed and the target's motion parameters are calculated. Compared with the traditional system based on PC, this system adapts increasing requirements of the target tracking system for image processing with real-time and bandwidth and has advantages of low cost, design flexibility, reconfigurable hardware, fast processing speed.Major work of this dissertation includes: the software and hardware platforms' constructing; Target detection algorithm is simulated, analyzed and compared in MATLAB; Timing design, function simulation and verification of functional modules are implemented using various EDA software tools such as Synplify Pro, Modelsim and TimingDesigner. Finally, the system engineering is synthesized and placement and routing is completed in Quartus II. Ensuring the correct timing in the report, system configuration file is downloaded to the FPGA development board.Implementation result shows that: this moving target tracking system works well in FPGA and meets with the design requirements. It provides the foundation of the visual intelligence surveillance.