Research On Infrared Moving Target Detection Algorithm And FPGA Hardware Implementation

With the development of infrared technology, infrared imaging equipment is widely used in military and civilian fields with advantages of remote working, concealed, day and night work, etc. However, in practical applications, detection environment is complex, size of infrared target is different, velocity of target is varied, how to keep a high integrity of target and high real-time of system in detection process is very important in the research of infrared target detection.This paper is divided into five parts:The first part introduces the significance and background of infrared moving target detection, and introduces the current research at home and abroad. The second part firstly introduces the characteristic of infrared image and the characteristic of infrared small target. Then this part expounds the basic principle of many infrared moving target detection algorithms, such as:Frame difference algorithm, LOBSTER algorithm and Gaussian mixture model algorithm. For the shortage of a relatively novel detection algorithm (VIBE algorithm), this paper proposed an improved visual background extraction algorithm. When the first frame of video exists moving target, the algorithm this paper presented can eliminates Ghost in the process of later detection. The third part introduces traditional performance evaluation index for moving target detection algorithm and presents a new evaluation index based on the principle of image block. This new evaluation index makes up for the deficiency of traditional one, and the calculation results are not only in accordance with visual perception of human, but also objectively evaluate the performance of the moving target detection algorithm. The fourth part introduces the design of target detection algorithm simulation interface. This interface can help people directly know the performance of moving target detection algorithm. The fifth part introduces the design of multispectral image acquisition and processing system, including the system works, design of front-end detection system and design of target detection system. Then this part introduces the FPGA programming about the improved visual background extraction algorithm. Experiments show the target detection system based on the improved visual background extraction algorithm can completely detect target with small error and high real-time, it meets the detection needs of the actual scene.