提出了一种基于传感器光谱特性的全色与多光谱影像融合方法.该方法主要基于快速IHS融合方法的思想,在充分考虑各波段成像传感器的相对光谱响应的基础上,对强度分量的构造和空间细节信息的注入方式2个方面进行了改进.改进后的方法有效地改善了光谱畸变问题,且能同时对所有的光谱波段进行融合.分别对IKONOS、ETM+影像进行融合实验,结果表明该方法在光谱信息损失最少的情况下,较大地提高了影像的空间分辨率,与其他方法相比具有更好的综合性能.

Formostof the traditional image fusionmethods, the spectraldistortion of the fused images is a common problem.The major reason relates closely to the different spectral characteristics of sensors. In this paper, a novel spectral preservation fusion method for remotely sensed images is presented by considering the spectralcharacteristics of sensors. Themethod ismainly based on the fast IHS transform. Afteranalyzing the spectral response of themultispectral and panchromatic sensors, the IHS fusionmethod has been improved in two parts: the construction of intensity component and the injection method of detail information. In the proposedmethod, the spectraldistortion is reduced significantly and all themultispectral bands can be fused at the same time. To evaluate the performance and efficiency of the proposedmethod, two experiments are carried out on IKONOS and ETM+images respectively. The results show thattheproposedmethod isbetter than othermethods. Itcankeep almostthe same spatialresolution as the panchromatic images andminimize colordistortion.