Target Extraction from the Military Infrared Image with Complex Texture Background

摘要

The texture characteristic is one of the important factors of infrared image. For detecting the region of interest, a method of target extraction from the infrared image with complex texture background was presented. First, Mean-shift smoothing algorithm was used to smooth the image pixels, and then an eight directions difference clustering process combined with Mean Shift segmentation was used to extract the region of target. The method is relatively simple making it easy for practical applications. Experimental results show that the method can extract the information of target from infrared images without surveillance and has better adaptability, indicating that it is an effective algorithm for military field with a certain practical value.