基于小波变换模极大值特征的多模医学图像融合算法研究

Fusion of Multimodality Medical Image Based on Characteristic of Wavelet Transform Modulus Maximum

文章提出了一种基于小波变换模极大值特征的多模医学图像融合算法,将待融合的图像进行小波变换提取其模极大值特征,在特征域上根据小波变换模极大值的特征实施不同权重的加权融合计算,实施逆变换重建融合图像。并对人脑的MRI-PET图像进行了融合,实验结果表明该方法能有效地将解剖信息和功能信息融合在一起,并保留原始图像的解剖结构特征。

In this paper,a multi-modality medical image fusion algorithm has been studied using the modulus maxima of images wavelet transform,the method is implemented based on the multiresolution characteristic of the coefficient which are derived from a wavelet transform modulus maxima.The scale-space behavior of the decomposition coefficient has been studying.A thresholding is performed dynamically according to similarity measure of the two image,examples are presented to demonstrate the efficiency of the technique on a fusion of brain MRI-T1and PET.