SAR与TM影像像元级融合方法探讨

The Fusion Method of TM and SAR Based on Pixels

以加拿大Radarsat SAR与美国Landsat TM影像为数据源,分别将SAR与TM的DN值转换为表征地物特征的后向散射系数和反射率,采用修正的SVR融合法进行融合,利用最佳指数法跟常用的融合方法作了定量比较.结果表明:修正后的SVR法融合后图像无论是光谱信息保持性、融合后信息量都优于常用的融合方法.

The Radarsat SAR image and Landsat TM image were selected as the materials, after converting the DN value to backscatter coefficient and reflectivity separately, the modified SVR method was used to fuse the images. It not only enhances spatial detail information of the fusioned image largely, but also preserves spectral information of the original multispectral image well.