Design And Implementation Of The Motion Target Tracking System Based On FPGA

The technology of motion target detection and tracking has been widely used in consumer electronics, industrial inspection, security surveillance and intelligent transportation. The processing speed of target detection and tracking has been a huge gap for system implementation. This system selected the FPGA as the platform, using its unique parallel processing mechanism and powerful computing to improve the system processing speed and performance.Nowadays, the research for moving target detection and tracking algorithm is so mature that the improving space is not large. The implement way that using C language in PC cant be satisfied with the performance requirements of processing speed. To seek higher processing speed, this paper put forward a method to design and implement system which is based on NIOS II. The bus clock of this system reached266.67MHz, combined with simple and effective detection and tracking algorithm, used hardware acceleration of floating point to make the processing speed improved greatly.This paper designs moving target tracking system through constructing hardware platform and software system, and divides it into module to realize. The system includes data input module, motion detection module and target tracking module and display module. The data module is realized through combined with SoPC Builder and NIOS II, transplants FAT file system, compiles the bottom of the preparation of SD card driver. Getting the parameters information of moving target, basic tracking algorithm’s characteristic is the target color, and then get video tracking with particle filter. In order to make the256MB DDR2SDRAM space to the cache for detection and tracking algorithm finally, this paper provide LTM IP core according to Avalon Interface Specifications. Then display the welcome interface and results through LTM.This system has the speed of20fps for processing true color images which is320*240. It realized the real-time process, and the tracking accuracy can reach more than93%. The results show that the system achieves the desired function, has better real-time and reliability, small volume, convenient control with a wide range of applications.