Abstract: The standard geometric active contour is a powerful segmentation method that has been wildely used for a long time now,yet there are some problems in practice .for example, it susceptible to weak edges and image noise. The paper propose a new geometric active contour that based on gradient flow forces with region constraints. The extra region force gives the model a global view of the boundary information within the image which helps detect fuzzy boundaries and overcome noisy regions.

Key words: geometric active contour, weak edge, image segementation

摘要： 传统的几何活动轮廓模型作为一种有效的图象分割方法，一直以来被广泛使用。但其在应用中也存在不少问题，例如对图象内弱边界区域目标不能准确的分割以及对噪声的干扰容易使曲线陷入局部极值等情况。本文提出了一种基于区域梯度流力的几何活动轮廓模型，由于区域力从全局的角度为模型提供目标边界信息，这样使该模型不但能够准确的利用区域信息分割出弱边界区域而且能够有效抵御噪声的干扰。

关键词: 几何活动轮廓, 弱边界, 图象分割