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

基于局部和全局灰度拟合的图像分割算法
引用本文:张晶.基于局部和全局灰度拟合的图像分割算法[J].科技信息,2010(35):I0218-I0220.
作者姓名:张晶
作者单位:琼台师范高等专科学校数理系,海南海口571100
摘    要:针对灰度分布非均匀图像的分割,提出一种改进的基于区域的活动轮廓模型,融合了LIF(local image fitting)模型的变尺度局部拟合特点与C-V(Chan-Vese)模型的全局优化特性,不仅提高了图像的分割效率,而且增强了模型对尺度参数和初始轮廓位置的鲁棒性。在数值计算中,使用高斯滤波规则水平集函数,使其保持光滑,并避免了复杂的重新初始化过程。对大脑MR图像的实验分割显示了该模型的优点。

关 键 词:图像分割  灰度不均匀  活动轮廓模型  水平集方法

An Improved Image Segmentation Methods Based on Local and Global Intensity Fitting Energy
ZHANG Jing.An Improved Image Segmentation Methods Based on Local and Global Intensity Fitting Energy[J].Science,2010(35):I0218-I0220.
Authors:ZHANG Jing
Institution:ZHANG Jing (Qiongtai Teachers College, Haikou Hainan, 571100, China)
Abstract:An improved region-based active contour model to segment images with intensity inhomogeneities is proposed. It combines the local intensity fitting at a controllable scale of LIF (local image fitting) model and the global optimization of C-V (Chan-Vcse) model. Not only the computational efficiency are improved, but also the Scale parameter and initialization of the contours can be flexible chosen: Moreover, Gaussian filtering is utilized to regularize the level set function, which can ensure its smoothness and eliminate the requirement of re-initialization. The experimental results show the advantages of the method applied to brain MR images in terms of computational efficiency and robustness.
Keywords:Image segmentation  Intensity inhomogcneity  Active contour model  Level set method
本文献已被 维普 等数据库收录!
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