摘要
提出一种基于小波变换和自组织特征映射(SOM)神经网络的医学图像融合方法,对图像进行小波变换,以图像的小波系数为特征,采用SOM网络对图像进行聚类,并进行模糊分类,从而确定像素融合的权重,得到融合图像。仿真实验结果表明,该方法能够获得良好的性能。
A medical image fusion method based on the wavelet transformation and Self-Organization feature Map（SOM） neural network is proposed in this paper. Wavelet transformation of image is done. The wavelet coefficients are used as the characteristics. The SOM network is used to realize the cluster of imagcs and the fuzzy classification is done. The image pixels fusion weights are determined according to the classified result. The fusion image is obtained. Simulation experimental results show this method can get better performance.
出处
《计算机工程》
CAS
CSCD
北大核心
2009年第21期200-202,205,共4页
Computer Engineering
基金
东北大学"985工程"项目"信息化基础结构关键技术科技创新平台"
关键词
图像融合
小波变换
自组织特征映射神经网络
聚类分析
image fusion
wavelet transform
Self-Organization feature Map（SOM） neural network
clustering analysis
