基于小波变换的图像融合算法  

An Image Fusion Algorithm Based on the Wavelet Transform

作　　者：宋凯[1] 纪建伟[1] 

机构地区：[1]沈阳农业大学信息与电气工程学院

出　　处：《沈阳农业大学学报》2007年第6期845-848,共4页Journal of Shenyang Agricultural University

基　　金：国家"863"计划项目(2003AA41250)

摘　　要：图像融合技术的研究目的就是综合不同类型的传感器所获取的图像信息,通过对多幅图像间冗余数据的处理提高图像的可靠性和可观察性,通过对多幅图像间互补信息的处理提高图像的效果。提出了一种基于小波的融合算法,该算法基于小波图像的能量集中在低频子带、细节体现在高频子带的特点,在低频部分对小波系数采用基于局部能量的加权融合方法处理,高频部分采用基于边缘信息的加权算法进行融合。结果表明:所采用的方法避免了图像融合过程中因平均化而出现的模糊现象,融合后的图像内容更加清晰,更容易识别。实践证明,基于小波变换的图像融合可以取得良好的结果和较快的处理速度。The main aim of the image fusion technology is to integrate image information which obtained by different types sen-sors,and improve the reliability and the observability of the image by processing the redundant data among the images. We can the effectiveress of image could be improved by processing the complementary information. A fusion algorithm based on the wavelet transform was proposed. This algorithm was based on the property that the energy of image lying in the lower frequency sub-band and the details lying in the higher frequency sub-band, the lower frequency band were fused by Weighting fusion algorithm based on energy,then the high frequency band were fused by Weighting fusion algorithm based on edge. Results showed that it could avoid the emergence of vagueness phenomenon because of image average, yield much more clear image and much easier to discern. The image fusion based on wavelet transform could be achieved better results and faster processing speed.

关 键 词：图像融合 小波变换 加权融合 MALLAT算法 低频部分 高频部分 

分 类 号：TN911.73[电子电信—通信与信息系统] TP751[电子电信—信息与通信工程]