首页 | 本学科首页   官方微博 | 高级检索  
     检索      

几种图像分割算法在CT图像分割上的实现和比较
引用本文:杨加,田捷.几种图像分割算法在CT图像分割上的实现和比较[J].北京理工大学学报,2000,20(6):720-724.
作者姓名:杨加  田捷
作者单位:1. 北京理工大学,电子工程系,北京,100081
2. 中国科学院,自动化研究所,北京,100080
基金项目:国家自然科学基金,国家高技术研究发展计划(863计划)
摘    要:对目前几种在图像分割领域得到较多应用的交互式分割、区域生长分割以及阈值分割算法进行了探讨,并且结合实际CT片图例分别进行分割实验研究,得到较为满意和可用性强的结果,实验表明:瓶值分割对于CT切片的效果最好;区域生长分割适宜于对面积不大的区域进行分割分割效果了;基于动态规划的交互式分割算法比较复杂,计算时间较长,但对于边缘较平滑的区域,同样具有较好的实际效果,几种算法的评估为其在CT图像分割上的实际应用提供了科学依据。

关 键 词:图像分割算法  CT图像分割  交互式分割  阈值分割

The Implementation and Evaluation of Several CT Image Segmentation Algorithms
YANG Ji,WU Qi yao,TIAN Jie,YANG Hua.The Implementation and Evaluation of Several CT Image Segmentation Algorithms[J].Journal of Beijing Institute of Technology(Natural Science Edition),2000,20(6):720-724.
Authors:YANG Ji  WU Qi yao  TIAN Jie  YANG Hua
Institution:YANG Jia 1,WU Qi yao 1,TIAN Jie 2,YANG Hua 2
Abstract:Interactive segmentation, region growing segmentation and thresholding segmentation are studied, which are used widely in image segmentation. These algorithms are applied in some experiments to deal with CT image and the satisfactory results are gotten. The results show that among these algorithms, thresholding segmentation does best in CT image segmentation; in dealing with relatively small images, satisfactory results can be got by using region growing segmentation; although more time will be needed in using interactive segmentation which is the most complex, it dose fairly well with images whose edges are relatively smooth. The evaluation of the three algorithms provides a scientific basis of the application of these algorithms.
Keywords:image segmentation algorithms  CT image segmentation  interactive segmentation  thresholding segmentation  region growing segmenta  t
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《北京理工大学学报》浏览原始摘要信息
点击此处可从《北京理工大学学报》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号