A method of small target detection based on energy accumulation and morphology in infrared image sequences

摘要

Aiming at the characteristic of infrared image whose background is complicate and noise is serious, this paper puts forward a fused detection algorithm based on energy accumulation and morphology. The energy accumulation can restrain the noise of Gauss distribution effectively. The morphological top-hat operator can remove background efficiently. Firstly, analyze the characteristics of small target, noise and background in infrared image and establish the mathematical model. Secondly, remove noise by energy accumulation. Thirdly, remove background by morphological top-hat operator. Finally, do binarization by global threshold, obtain target and a few of remainder by opening and closing operations, so the remainder is removed commendably, and the resumed result is close to target. It has been proven by the simulation results that the method is rational and efficient.