SEGMENTATION OF INFRARED WEAK AND SMALL TARGET IMAGE BASED ON CELLULAR AUTOMATA
基于元胞自动机的红外弱小目标图像分割

For segmenting weak and small target image in infrared imaging guidance systems, a new automatic segmentation method based on cellular automata was proposed. First, the image contrast was enhanced by using cellular automata, and a von Neumann neighborhood with a uniform cellular automaton rule was adopted for the state transition function. Then, the image was binarized by using the method of dual edge thresholds. Finally, the target blocks were labeled and filtered according to seed points and target complexity. In this way the accurate results of image segmenting were obtained. In the segmentation experiments for three different types of infrared images captured in the outfield, the proposed method can locate effectively the region of multiple weak and small targets, which facilitates the subsequent target recognition and tracking.