基于二代Bandelet和主成分变换的高光谱遥感图像融合

Hyperspectral Remote Sensing Images Fusion Algorithm Based on Second Generation Bandelet and PCA Transform

摘要:

针对高光谱遥感图像具有波段多、波段间冗余大的特点,提出一种基于二代Bandelet和主成分(principal components analysis,PCA)变换的高光谱遥感图像融合的方法,利用二代Bandelet变换进行图像的多尺度几何分析,得到每个波段图像的Bandelet系数和几何流,对多个波段Bandelet系数和几何流进行PCA变换,得到其主成分,逆变换重构图像.实验结果表明，基于二代Bandelet和PCA变换的方法能很好地融合高光谱遥感图像,优于传统的Bandelet变换和PCA变换.

Abstract:

According to the characteristics of hyperspectral remote sensing images such as multiband and data redundancy,a novel fusion method of hyperspectral remote sensing images based on the second generation Bandelet and PCA transform was proposed.Bandelet transform was performed to gain Bandelet coefficients and geometries of subbands.Then PCA transform was performed to calculate their principal component.Finally,fused images were reconstructed with their principal component by taking inverse Bandelet transform.The experiment shows that this method can well merge hyperspectral remote sensing images.Its result is better than those of Bandelet and PCA transform method.