基于粗糙模糊聚类的脑MR图像分割技术
关键词 磁共振图像 图像分割 粗糙集 模糊聚类 噪声 偏移场 灰度不均匀
毕业论文设计说明书外文摘要
Title A Rough-fuzzy C-Means Clustering Algorithm for Magnetic Resonance Image Segmentation
Abstract
Magnetic resonance imaging technology given its advantages and practicality has become an important adjunct to the clinical diagnosis of brain diseases. Accurate segmentation of brain MR images is quite important for biomedical research and clinical applications. However, in practice, there are a few defects including intensity inhomogeneity of brain MR image, noise, partial volume effects of brain tissue and low-contrast,so that accurate segmentation of brain MR images becomes very difficult. In this paper, in order to solve segmentation problem of the Brain MR image, an effective brain fuzzy C-means clustering rough segmentation model based on nuclear magnetic resonance is put forward. The model is pided approximation image set by using an internal data structure, this way can effectively overcome the algorithm for the initialization parameter sensitivity; the introduction of polynomial basis functions can overcome the effects of gray unevenness . Experimental results show that this model has a strong anti-noise ,get accurate estimation image segmentation results of bias field at the same time ,and has strong robustness for parameter initialization.
Keywords brain magnetic resonance image segmentation; rough sets; fuzzy clustering model; noise;bias field; intensity inhomogeneity
目 次
1 绪论 1
1.1 研究背景 1
1.2 脑MR分割难点 1
1.3 国内外研究现状分析与发展趋势 2
2 粗糙集以及模糊聚类模型 4
2.1 聚类分析 4
2.2 粗糙集 5
2.3 粗糙模糊C均值算法 5
3 含有噪声以及偏移场的MR图像的相关分割方法和策略 6
3.1 含噪声脑MR图像的分割方法和策略 6
3.1.1 策略一：先去噪后分割 6
3.1.2 策略二：分割过程中克服噪声 7
3.2 含偏移场脑MR图像的分割方法和策略 7
3.2.1 策略三：先估计偏移场后分割 8
3.2.2 策略四：在分割过程中估计偏移场 8
4 粗糙模糊聚类的MR图像分割 10
4.1 解决的缺陷 10
4.2 粗糙模糊C均值混合模型（Rough-Fuzzy C-means,RFCM） 10
4.3 基于粗糙模糊集聚类的脑MR图像分割模型（Generalized Rough-Fuzzy C-Means, GRFCM） 11
上一篇：基于日志的事件摘要模块设计与实现
下一篇：汽车虚拟中控平台的仿真与软件实现