关键词: 局部熵, 主动轮廓模型, LGDF 模型, 图像分割, 水平集

Abstract:

An active contour model, which combines LGDF model and CV model and uses the image entropy, is proposed to solve the problem of sensitivity to the initial contour of the original active contour. First the level set force is defined by a linear combination of the LGDF and CV models incorporating image entropy. Then the weights of the forces are regulated by the local image information. Experiments on blood X-images, noise images, and SAR image show that the proposed method is robust to initialization and less sensitive to noises.

Key words: local entropy, active contour, local gaussian distribution fitting, image segmentation, level set