Contourlet变换系数加权的医学图像融合  

Medical Image Fusion Based on Contourlet Transformation

作　　者：王昌[1] 刘艳[1] 张文超[1] 

机构地区：[1]新乡医学院生物医学工程学院,河南新乡453003

出　　处：《科技视界》2015年第30期34-34,60,共2页Science & Technology Vision

基　　金：新乡医学院科研培育基金资助项目(2014QN142)

摘　　要：本文针对多模态的CT、MRI图像的融合,利用Contourlet变换多尺度、多方向的特性,提出了系数加权的融合算法。对于低频子带,采用局部信息均值和方差来进行系数加权;对于高频子带,采用条件加权。将本算法应用于不同模态的医学图像的融合,有效增加多模态医学图像的互补信息,提高融合的清晰度。In order to solve the problem of multi-modality image fusion, an image fusion algorithm based on Contourlet transform was proposed in this paper, using the multi scale and direction characteristics of Contourlet transformation. For low frequency sub bands, the mean and variance of the local information is used to carry on the weighted coefficients; for the high frequency sub bands, the e.onditional weighted is used. This algorithm is applied to the fusion of medical images of different modalities, and the complementary information of multi modulity medical images can be effectively improved.

关 键 词：多模态 医学图像 CONTOURLET变换 图像融合 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]