山东理工大学学报（自然科学版）2018，Vol.32Issue(3)：13-16,20,5.
基于Nystrom方法的水平集医学图像分割算法
Level set method of medical image segmentation algorithm based on Nystrom method
摘要
Abstract
The existing image segmentation model based on minimized region-extended fitting en-ergy has a tendency to produce edge leakage for edges with fuzzy and strong noise,resulting in dissatisfactory segmentation.A new level set of medical image segmentation algorithm based on Nystrom method is proposed for this phenomenon.Firstly,the original image is sampled by the Nystrom method,and the similarity matrix and eigenvector are estimated approximately.Then the feature vector is clustered by the k-means algorithm.Finally,the image segmentation is real-ized by the level set segmentation method.The experimental results show that the segmentation time is reduced and the similarity coefficient is improved with the same number of iterations com-pared with the minimized region-extended fitting energy for image segmentation.关键词
水平集/图像分割/聚类/Nystrom方法/相似矩阵Key words
level set/image segmentation/clustering/Nystrom method/similarity matrix分类
信息技术与安全科学引用本文复制引用
王晓丹,张龙波,王雷,刘晨..基于Nystrom方法的水平集医学图像分割算法[J].山东理工大学学报（自然科学版）,2018,32(3):13-16,20,5.基金项目
国家自然科学基金青年科学基金项目(61502282) （61502282）
山东省自然科学基金青年科学基金项目(ZR2015FQ005) （ZR2015FQ005）
山东理工大学博士科研启动经费资助项目(414023) （414023）
淄博市校城融合发展计划项目(2016ZBXC049) （2016ZBXC049）
