一种基于模糊集和小波变换的图像融合方法  

Image fusion scheme based on fuzzy sets and wavelet transform

作　　者：李涛[1] 田岩[2] 柳健[2] 唐东峰[3] 

机构地区：[1]华中科技大学图像识别与人工智能研究所,湖北武汉430074 [2]华中科技大学电子与信息工程系,湖北武汉430074 [3]湖南科技大学信息与电气工程学院,湖南湘潭411201

出　　处：《华中科技大学学报（自然科学版）》2005年第3期24-26,共3页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目 (6 0 2 72 0 99) ;测绘遥感信息工程国家重点实验室开放基金资助项目 .

摘　　要：提出了一种基于小波变换和模糊集的图像融合算法 .其基本思想是 :首先对图像进行小波变换 ,获得图像的低频和高频分量 ;随后在融合过程中 ,对低频和高频分量采取不同的融合策略 ,即对低频分量采用平均能量法进行融合 ,对高频成分利用图像的模糊集 ,寻求一个模糊隶属函数作为融合算子进行融合 ;最后再对融合后的图像进行小波反变换 ,重构出融合后的图像 .实验结果证明了方法的有效性 .A novel image fusion scheme was presented based on wavelet transform and fuzzy sets. Two images which have been registered were first cutted up into one low and three frequency components using wavelet transformation. Then during the image fusion, different rules were used for different components. Average energy was used as the fusion operator for the low frequency part. The coefficients of the high frequency parts were been mapped into fuzzy sets respectively. The belonging function was used as the fusion operator for the high frequency parts. The fused image was obtained by taking inverse wavelet transform. The experimental results show that fusion scheme is effective.

关 键 词：图像融合 模糊算法 隶属函数 小波变换 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]