基于小波的影像融合方法研究  

Research of Image Fusion Methods Based on Wavelet Theory

作　　者：李敬[1] 蔡晓斌[2] 周新志[1] 

机构地区：[1]四川大学电子信息学院,四川成都610064 [2]武汉大学测绘遥感信息工程国家重点实验室,湖北武汉430079

出　　处：《现代电子技术》2006年第11期42-45,共4页Modern Electronics Technique

摘　　要：影像融合技术作为一种数据整合技术,能够整合不同影像中高空间分辨率和高光谱分辨率信息,因其能明显改善影像的显示效果,因而被广泛的应用于遥感图像处理中。传统的HIS变换、主成分分析、Brovey变换虽然对于影像的质量能起到一定改善的作用,但由于其自身的算法的原因,还存在许多不足。采用小波HIS、小波PCA变换等方法进行了新的融合方法的有益尝试。以IKONOS 2的1 m和4 m分辨率影像数据为实验数据,对比了小波融合方法与传统融合方法,实验结果显示小波方法在定性和定量方面都有明显优势。文中还讨论了对融合后影像质量的多种定量评价方法。As a data integration method,the image fusion can get all advantages from the high - spatial resolution and high - spectral resolution images. Because the method can improve the image quality greatly, so it is used widely in the remote sensing image processing. For the shortage of algorithms, some custom fusion method couldn＇t solve the flaw of fused image. For getting the better fusion result, it induces the wavelet theory to the custom method and applies the method in the fusion processing between IKONOS Pan image and multi - spectral image. Finally the quantity and quality comparison between the wavelet fusion results and the custom fusion results are fulfilled. It shows that the wavelet fusion methods can get better effect. In the same time,many quantity assessment parameters are introduced and used to evaluate the fusion result.

关 键 词：IKONOS 影像融合 小波HIS变换 小波PCA变换 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]