Research On Visual Simulation And Detection And Tracking Technology Of Infrared Moving Target

Infrared moving target tracking technology has become an important aspect of target tracking in military and civilian fields, but most tracking technology can’t meet the real-time requirement of video processing, and many tracking algorithms need to select tracked target manually. With the performance advantage of FPGA in video processing, in this paper, we propose an infrared target detection and tracking algorithm based on visual attention and mean shift.The main content of this paper is divided into four parts. The first part introduces the infrared image pre-processing technology, expounding the basic principle of many infrared image processing algorithms, such as image smoothing, morphological processing, image threshold segmentation. The second part introduces the visual simulation technology of infrared moving targets, we develop an infrared scene simulation software based on MFC and Vega. It can set different scenes and atmospheric conditions as we need, what’s more, it can add moving targets and change environmental factors dynamically. The third part describes the proposed infrared moving target detection and tracking algorithm based on visual attention and mean shift. Firstly, infrared video sequence will be processed by visual attention algorithm. Secondly, moving object will be extracted after thresholding. Thirdly, the algorithm will call improved mean shift algorithm to track the moving target. Then we simulate the algorithm by Matlab software. The fourth part introduces the FPGA-based infrared video capture and processing system, we complete the design of visual attention algorithm based on hardware in the Quartus II software platform. The experimental result shows that the algorithm has advantages of quick processing speed and it doesn’t need to select target manually when compared with common tracking algorithm.