为了有效去除遥感影像中数据冗余问题,提出了一种基于受限的非负矩阵分解的影像融合算法,利用矩阵的非负性实现了中巴卫星多光谱影像和Landsat ETM+高分辨率全色影像的融合。实验结果表明,与IHS变换,小波变换等传统融合方法比较,该方法在较好的保留光谱信息的同时,空间细节信息也得到了增强,同时具有较高的峰值信噪比。
Data fusion on remote sensing is an important problem in image processing. The key for a successful image fusion is to find an effective and practical image fusion algorithm. To eliminate data redundancy for two different remote sensing images, a new approach using constrained non-negative matrix factorization for remote image fusion between Landsat ETM+panchromatic and CBERS multi-spectral images was proposed. Visual and statistical analyses proved that the concept of fusion based on constrained non- negative matrix factorization was promising, and it improved significantly fusion quality and signal-to-noise ratio when compared to conventional IHS and wavelet fusion techniques.