Abstract: The segmentation algorithm for setting the gate-limit which is commonly used for image processing can segment targets and backgrounds effectively. But the suitable adjusting coefficient must be chosen in this method. Aiming at the question, a novel algorithm for adaptive image segmentation of moving targets is presented. Based on setting self-tracking widows for tracking moving targets, the adjustable coefficient can be found adaptively on the Otsu evaluation function, and then adaptive image segmentation is completed by setting the gate-limit of image segmentation. The novel method is tested, and compared with the Otsu method, the iteration method and the maximum entropy method. Results show that the algorithm is not only valid for various complex backgrounds, but also has the features of high precision and fast speed, and that it is a practicable and effective image segmentation method.

Key words: image segmentation, adaptive algorithm, Otsu evaluation function, moving targets