Abstract：This paper analyses the gray characteristic of the infrared dim target under complex background, and proposes a new target detection method. Firstly, an improved detection template is introduced based on the analysis of the gray distribution characteristic in different structure of an infrared image. The adopted detection template has eight different directions and three levels, which can filter out some parts belonging to the background. Secondly, a pretreatment process based on mean filter is adopted to save detection time consumption and a new improved algorithm based on gray characteristic is proposed. Finally, experimental results reveal that the proposed method can effectively detect small and dim targets, which builds a basis for improving the detection and tracking efficiency of missiles.
