基于整数小波和非负有序独立成分分析的遥感影像融合

Remote sensing image fusion based on IWT-ONICA

College of Information Science and Technology, Beijing Normal University, Beijing 100875, China;2. Institute of Geology and Geophysics ,Chinese Academy of Sciences, Beijing 100029, China

摘要: 结合ICA算法可以去除高阶冗余信息和IWT具有多分辨率、运算速度快的特性，在非负性条件的约束下，提出一种基于整数小波变换和非负有序独立成分分析(IWT-ONICA)的遥感影像融合方法。将该方法用于全色影像和多光谱影像的融合，数理统计(峰值信噪比、扭曲度、熵等)和目视判读结果表明，该方法融合效果优于其他方法，融合结果更加清晰，可视性更好。
- 整数小波变换 /
- 非负有序独立成分分析 /
- 遥感影像融合
Abstract: A novel remote sensing image fusion method based on integer wavelet transformation and ordered non-negative independent component analysis (IWT-ONICA) is proposed according to the characteristic that ICA can remove the higher-order redundancy. The processing result of IWT is multi-resolution and the computing speed of IWT is quick. The proposed method is used to the fusion of Landsat ETM+ and CBERS images. Mathematical statistical and visual results show that IWT-ONICA is superior to other fusion algorithm. The fused image of IWT-ONICA is clearer and the visual effect is better.
- integer wavelet transformation /
- non-negative ordered ICA /
- remote sensing image fusion