基于活动轮廓模型的左侧脑室图像分割

Image Segmentation for Left Lateral Ventricle Based on Active Contour Model

在高斯力活动轮廓模型基础上,提出了径向力活动轮廓模型,可增加外力的作用范围,扩大初始轮廓曲线收敛区域。实验结果表明,该模型能正确分割左侧脑室图像,并且初始轮廓曲线愈接近目标真实边界,迭代次数愈少,分割速度愈快。

A radial force active contour model for image segmentation is proposed based on Gaussian force active contour model in the paper. This new model has enhanced sphere of influence in force field and enlarged convergence domain of initial contour curve. The experimental results have showed that left lateral ventricle is accurately segmented with the model, and the nearer to objective boundary initial contour curve, is the less iteration number, and the higher segmentation speed.