Dsp Image Tracking System Hardware And Software Research

Target identification and tracking have important application in military and civilian. For decades, Identification and tracking of targets under different backgrounds have always been an important research area. With the development of digital-signal-processing technology, it is become more and more simple and effective to solve this problem with high speed DSP chip which has powerful computing.In this dissertation, the design of target detection and recognition system based on high-speed DSP is studied, aimed mainly to solve the target's tracking problem in low signal-to-noise ratio circumstance.Based on the deeply research of the newest development in video tracking and DSP technology,an intensive study of some key problems in the design of the high speed DSP video tracking system is made.For the requirement of anti-shade of the low signal-noise-ration target, The tracking algorithm integrating background subtraction and block matching is employed, meantime, Kalman filter which used to predict the block matching point is verified through the experiment. Considering the development trend of digital signal processor (DSP), TI's latest video processing DaVinci DSP——DM6446 is choosed as the core to build the system hardware platform.Also designed key module circuits include video input, storage, processing, video output, communications and motion control etc. Based on the TI's integrated development environment (CCS),the video driver,target detection,image tracking and tilt control algorithm and the corresponding software are designed. All software are verified on the high-speed DSP image processing platform.