摘要
距骨的图像分割是对其进行三维可视化及运动学研究的基础。提出一种距骨的交互式图像分割算法,由三部分组成:(1)提出一种改进的迭代阈值法,从一组足部CT图像序列中分割出所有骨骼。(2)提出一种加速的live-wire分割算法,断开距骨与其他骨骼的连接线,该方法与原Live-wire算法相比,代价计算量大为减小且缩短了最优路径搜索时间。(3)用形态学操作提取距骨。实验结果表明,该算法能以较少的交互次数快速地进行距骨图像分割,也可用于其他骨骼的分割。
Image segmentation of talus is the foundation of its three dimension visualisation and kinematics study. An interactive image segmentation algorithm of talus is presented, which is composed of three parts ： （ 1 ） An improved iterative threshold method is proposed to separate all of the bones from a set of CT image series of foot. （ 2 ） An accelerated Live - wire segmentation algorithm is presented to disconnect the connecting line between the talus and other bones, this method reduces the computation of cost greatly and shortens the optimal path searching time compared with original Live -wire algorithm. （3） The talus is extracted using morphological operation. Results show that this algorithm can quickly segment talus image with fewer interactive times. It can also be used for other bones segmentation.
出处
《计算机应用与软件》
CSCD
2009年第5期77-80,共4页
Computer Applications and Software
基金
上海市教育委员会重点学科项目(J51801)
关键词
距骨
图像分割
迭代阈值法
加速的Live-wire算法
形态学操作
Talus Image segmentation Iterative threshold method Accelerated live-wire segmentation algorithm Morphological operations
