Based On The Dm642 Infrared Moving Target Detection And Tracking Technology Implemented

Infrared imaging technology as one of the hottest technology currently has been widely used in Video Surveillance, Infrared Precision Guidance, Warning System, Searching and Tracking. It mainly depends on receiving the infrared radiation from target, with the advantages of all-weather working condition, excellent concealing characteristic and so on. With the continuous development of infrared imaging, infrared imaging of the motion target detection and tracking system came into being, and become an important technology in these areas, it has a very important position and broad application prospects in military and civilian areas.This paper focuses on the infrared moving object detection and tracking algorithm, and discusses the simulation of the infrared target detection and tracking, translates the model to C code of the target processor, and verify the correctness of the code conversion. First, description the various target detection and tracking algorithm theory detailed, comparing advantages and disadvantages, then, according to actual needs and the comparison results of those algorithms, some of them are optimized in combination with the image pre-processing. On this basis, then use Simulink model to achieving three kinds of detecting algorithms and two kinds of tracking algorithms. These three detecting algorithm are Temporal differencing, Background subtraction, Optical Flow. And these two tracking algorithms are Kalman, Artificial Neural Network. Then converse the Simuink models to C code base on DM642 processor automaticly, using PIL to verify the correctness of the code. This paper is very significant in the application of the infrared target detecting and tracking technology. And it has some value for the pre-research work in the project of "Ultra-low Altitude Target Detecting, Identification and Tracking Technology".