Active contours based algorithm for infrared image segmentation
红外图像活动轮廓分割算法

Infrared images always have blurry edges, low contrast and complicated background, therefore the segmentation of infrared images by traditional active contours is very difficult. This paper proposed an edge-based active contour model for the segmentation of infrared images. This model used a new edge stopping function which used the local entropy of image. To further improve the efficiency, this model used an automatic initialization scheme which was based on the feature of infrared image. The experiments show that the proposed model can efficiently segment infrared images.