Design of real-time small target detection system for infrared image based on FPGA

A simple detection measurement is developed for real-time small target detection in infrared background, which can be easily available on FPGA. Basing on the difference between target and background, a predictive background is generated via restraining the infrared background with median filter, and then global threshold segmentation is used to residual image for further processing. Taking advantage of the relationship of targets in frames, a comprehensive analysis of residual image is made, and then the precise position of target is got by removing pseudo-objectives. Experiment results show that the method is simple to apply to hardware system, improve the system´s false alarm rate and meet the requirements of practical engineering.