基于局部能量的Trous小波和IHS变换的影像融合研究

A study on image fusion by Trous wavelet and IHS translation based on Local energy rule

传统基于小波变换的影像融合随着分解级数的增加,融合影像与原多光谱影像间的相关性随之降低,融合后影像的小目标对象很难获得丰富的颜色信息。鉴于此,本文提出基于局部能量的Trous小波和IHS变换相结合的融合算法,对IKONOS影像进行融合,并与传统的Trous小波变换融合、Trous小波变换和IHS变换结合的融合算法相比较,结果表明改进的方法能提高融合影像的相关性,降低光谱扭曲度,增强小目标的识别能力。

With the increase level of traditional wavelet decomposition, the correlation of fused image with original multispectral image is lower inevitably, and even small objects are difficult to obtain color information.With the emergence of high revolution images such as IKONOS, the traditional methods are not satisfied for higher interpretation required.Based on this phenomenon, this paper presents a new method that combines local energy fusion rule with traditional Trous and IHS integrated algorithm to fuse the IKONOS image, and compares it with traditional Trous algorithm as well as traditional Trous and IHS integrated algorithm.The results indicate that the new method can increase the correlation and decrease the spectral warping degree on greater intensity, and also improve the recognition capacity for small objects.