关键词: CT图像, 模糊C均值, 轮廓跟踪, 三维模型重建

Abstract: Brain CT images always have some redundancy such as metal slots，clothes and shoulders.This paper divides the redundancy into two kinds based on its distribution and proposes a new algorithm which combines the FCM and contour track to clear redundancy in CT images.The algorithm solves the problem of the lost of CT data in traditional method.The algorithm can automatically deal with a batch of CT data and it is proved to have good effect and some apply value through experiments.

Key words: Computed Tomography（CT） image, fuzzy-C means clustering, contour track, 3D model reconstruction