| 基于高斯混合模型的人脑MR图像分割新算法研究 |
| |
| 引用本文： | 朱泉同,张建伟,陈允杰.基于高斯混合模型的人脑MR图像分割新算法研究[J].计算机应用与软件,2009,26(3). |
| |
| 作者姓名： | 朱泉同 张建伟 陈允杰 |
| |
| 作者单位： | 1. 南京信息工程大学数理学院,江苏,南京,210044
2. 南京信息工程大学数理学院,江苏,南京,210044南京理工大学计算机科学与技术学院,江苏,南京,210094 |
| |
| 摘 要： | 有限高斯混合模型是广泛应用于聚类分析与分布估计的概率模型之一,同样在脑部MR图像分割领域获得了广泛应用.利用高斯混合模型可以描述大脑图像,通过期望最大算法求解随机变量的特征值,并用其对图像上的点进行分类,可以在一定程度上解决脑图像分割问题.针对含脉冲噪声的大脑图像,首先利用改进的滤波方法对图像进行滤波,再利用粒子群改进算法的全局优化特性求解高斯混合模型的参数,这样避免了EM算法易陷入局部极值的现象,以提高参数精度,从而进一步提高分割质量.
|
| 关 键 词： | 图像分割 高斯混合模型 粒子群优化算法 EM算法 |
ON NEW SEGMENTATION ALGORITHM OF BRAIN MAGNETIC RESONANCE IMAGES BASED ON GAUSSIAN MIXTURE MODELS |
| |
| ZHU Quantong,ZHANG Jianwei,CHEN Yunjie.ON NEW SEGMENTATION ALGORITHM OF BRAIN MAGNETIC RESONANCE IMAGES BASED ON GAUSSIAN MIXTURE MODELS[J].Computer Applications and Software,2009,26(3). |
| |
| Authors: | ZHU Quantong ZHANG Jianwei CHEN Yunjie |
| |
| Affiliation: | College of Mathematics and Physics;Nanjing University of Information Science and Technology;Nanjing 210044;Jiangsu;China;School of Computer Science and Technology;Nanjing University of Science and Technology;Nanjing 210094;China |
| |
| Abstract: | The finite Gaussian Mixture model is one of the probability models,which is widely used in clustering analyses and distributing estimation.It has been widely used in image segmentation of brain images as well.A brain image could be described by Gaussian Mixture model,EM(Expected Maximum) algorithm is used to calculate the eigenvalue of random variant,and to classify the points on the image,therefore,the brain image segmentation issue could be resolved to some extent.In the paper,in light of the brain image ... |
| |
| Keywords: | Image segmentation Gaussian mixture models Particle swarm optimization Expectation maximization algorithm |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
