基于多小波变换的医学图像融合方法研究  

Methods Study of Medical Image Fusion Based on Multiwavelet Transform

作　　者：孙宏伟[1] 周振环[1] 

机构地区：[1]深圳职业技术学院电信学院,深圳518055

出　　处：《计算机工程与应用》2006年第23期215-219,228,共6页Computer Engineering and Applications

基　　金：深圳市科技计划项目(编号:05KJcd004);粤港关键领域重点项目(编号:2005A11304003)

摘　　要：医学图像融合在临床诊断及治疗上有重要意义,多小波是小波理论的扩展,在图像处理方面具有单小波所不具有的优点。文章对基于多小波变换的医学图像融合方法进行了研究总结,提出了基于边沿信息方向性加权的区域能量的融合规则,通过对CT与MRI图、CT与SPECT图实验结果比较,表明该文提出的算法具有更好的融合效果。Medical image fusion is significant in practical clinic application and therapy.Multiwavelet is an extension from wavelet theory,and has several particular advantages in comparison with scalar wavelets on image processing.In this paper,there is research report of medical image fusion based on multiwavelet transform.A new technique of the area energy image fusion rule of weighted direction with edge information is developed based on multiwavelet transform.Compared fusion results of CT and MRI images,CT and SPECT images are given,which show our proposed method can achieve better fusion performance than other usual fusion rules.

关 键 词：多小波变换 医学图像 图像融合 方向加权 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]