Adaptive Target Detection Based on Improved Genetic Algorithm in Infrared Images

A novel background estimation algorithm is proposed for target detection in infrared images,and we employ it in sky background small target detection and complex ground background human target detection application.The experimental results show that with the assistant of the trained structure elements and morphology target detection method,the complex background can be removed properly;the noise also can be effectively restrained,and the algorithm's accuracy and false positives both are improved.