IKONOS全色与多光谱数据融合方法的比较研究

IKNOS-2给各个应用领域提供1 m的全色和4 m的多光谱数据,因此利用全色波段将4 m的多光谱数据融合为1 m的多光谱数据会充分利用二者的信息,提高目视和自动影像提取的类别精度。影像融合技术发展较快,成为遥感应用研究领域的重要主题。在遥感领域应用较多的融合方法有IHS变换、主成分分析、Brovey(颜色归一化)变换、小波变换以及最近发展修改的合成变量比值变换。从光谱质量和空间信息角度对融合方法进行了比较研究,筛选出适合IKNOS融合方法,从各种特征信息提取和自动分类角度出发,合成变量比值变换融合方法光谱退化最小,同时也较高程度地保持了高几何分辨率全色的空间信息。

Study on Fusion Algorithms of IKONOS PAN and Multi-spectral Images

IKONOS image provides the 1m high spatial pan and the 4 m multi\|spectral data, which can be merged to improve the class accuracy of visual or computer interpretation. The fusion techniques have been developed quickly in recent years, an important remote sensing research topic. The fusion algorithms which are often used in remote sensing includes the IHS transformation, principal component transformation, Brovey transformation, wavelet transformation and the synthetic variable ratio modified recently. But it has less work on the evaluation of the quality of the different methods. This paper compares the different merging methods from the spectral quality and the spatial quality in order to choose the suitable method for the IKONOS image. From the view of character transform and computer classification, the synthetic variable ratio is the best technique which distort the spectral quality more slightly and combine the high spatial information better at the same time.