Nonnegative matrix factorization-based hyperspectral and panchromatic image fusion

摘要: The fusion of hyperspectral image and panchromatic is an effective process to obtain with both high spatial spectral resolutions. However, the property stored in original often distorted when using class traditional techniques. Therefore, this paper, we show how explicitly incorporating notion “spectra preservation” improve resolution fused image. First, a new model, preservation based on nonnegative matrix factorization (SPNMF), developed. Additionally, multiplicative algorithm aiming at get numerical solution proposed model presented. Finally, experiments synthetic real data demonstrate SPNMF superior technique for it could resolutions images their properties reliably preserved.