摘要
医学图像处理中,目标提取在准确区分组织结构中起了重要的作用,论文介绍了基于最大流的交互式的目标提取算法,该方法把基于区域和基于边界方法结合起来,并用分水岭算法对图像做预分割,在其所分的区域上计算能量函数,然后最小化能量函数求最优边界。笔者结合自己研究的课题,针对目标标记步骤中聚类种子F和#色彩值的k-means算法进行改进,并把改进算法应用到人体器官切片目标分割中。
Object extraction plays an important role in distinguishing organize framework accurately in medical image process.This text has introduced an interactive object extraction on the basis of max-flow,this method combines the basis of area with the basis of border method,and using watershed algorithm for image pre-segmentation,then calculating energy function on the segmented regions from it.The author uses one's own subject for research,applies this kind of algorithm to the human organ seg- mentation and improves in the k-means method by which the colors in seeds F and β are clustered in object marking.
出处
《计算机工程与应用》
CSCD
北大核心
2007年第5期246-248,共3页
Computer Engineering and Applications
基金
国家"十五"科技公关计划项目(the Key Technologies R&D Program of the "Tenth Five-Year-Plan" of China under Grant No.2001BA609A-5)
关键词
目标提取
最大流
图切割
object extraction
Max-flow
graph cut