基于多小波变换的遥感图像模糊融合算法研究  

Fuzzy Fusion Algorithm of Remote Sensing Images Based on Multi-wavelet Transform

作　　者：孙娇娇[1] 徐礼鹏[1] 邵雪松[2] 刘启蒙[1] 

机构地区：[1]安徽理工大学测绘学院,安徽淮南232001 [2]中国科学院自动化研究所,北京100190

出　　处：《地矿测绘》2011年第3期1-4,共4页Surveying and Mapping of Geology and Mineral Resources

摘　　要：基于多小波变换方法,提出了一种针对多光谱图像和全色图像的像素级遥感图像模糊推理融合算法。该算法首先利用IHS变换对多光谱图像进行颜色空间转换,然后对I分量和灰度拉伸后的全色图像进行多小波分解分别得到高频部分和低频部分图像。针对高频图像抗噪能力差的问题,提出了一种基于模糊推理的加权融合规则。对融合后的图像进行多小波重构和IHS反变换得到融合后的目标遥感图像。这种融合算法既保留了多小波变换在图像融合中的优点,同时也克服了传统融合规则会引起的图像模糊、噪声敏感的问题。实验结果表明,该方法在增加图像信息、增强图像目标特性和提高图像清晰度等方面有较好的融合效果。Based on the method of multi-wavelet transform, this thesis proposes the algorithm of the pixel level of a fuzzy reasoning fusion aiming at multi-spectral image and full color image. Firstly, the multi-spectral image is converted to IHS space through IHS transform, and then, with multi-wavelet decomposition, the intensity component of the multispectral image and the panchromatic image are transformed into high frequency parts and low frequency parts respectively. The weighted average fusion rule according to the fuzzy reasoning is investigated to solve the problem of high sensitivi- ty to noise of high frequency images. The final fusion image can be obtained through multi-wavelet reconstruction and inverse IHS transform. The fuzzy fusion algorithm not only keeps the advantages of multi-wavelet transform for image fusion techniques but also overcomes the shortcomings and noise susceptibility while employing traditional rules. The analysis of experimental results demonstrates the effectiveness of the proposed algorithm in the aspects of increasing image information, reinforcing the target characteristics and improving the definition.

关 键 词：多小波变换 模糊融合 IHS变换 遥感图像 

分 类 号：P237[天文地球—摄影测量与遥感] TP751[天文地球—测绘科学与技术]