基于小波加权系数和平均梯度的医学图像融合  

Medical Image Fusion Based on Wavelet Weight Coefficients and Average Gradient

作　　者：任青[1] 

机构地区：[1]山西农业大学信息学院,太谷030800

出　　处：《计算机与数字工程》2016年第9期1687-1688,1704,共3页Computer & Digital Engineering

基　　金：山西省自然基金(编号:2007012003)资助

摘　　要：针对医学CT图像和MRI图像的多模图像融合问题,论文提出了一种基于小波变换的改进算法。对于图像的低频部分,将小波低频系数乘以加权因子1/R,减少低频部分所占整个图像信息比例;对于图像的高频部分,利用图像质量评价中的评价因子——平均梯度绝对值取大来选取最终融合图像的小波高频系数。实验结果表明,该方法得到的融合图像体现出更强的融合性能。This paper proposes an medical image fusion algorithm,based on wavelet transform,it is mainly concerned with the CT and MRI.For the image low-frequency part,because most of the image characteristics are stored in high-frequency part,so low-frequency wavelet coefficients will multiply coefficien 1/R,to reduce the proportion of low-frequency part of the whole image information.For the image high-frequency part,evaluation factor of image quality evaluation—the average gradient,whose absolute value is larger is chosen to play as the ultimate image wavelet high-frequency coefficients.Compared with several commonly used image fusion methods,the proposed algorithm can give us the fusion image that ownes more image detail information.

关 键 词：图像融合 多模图像 小波变换 加权系数 平均梯度 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]