Super Resolution Surface Reflectance Generation
through Multi-Satellite Image Fusion

Recently, satellite data has become an essential foundational resource in various scientific fields. As a result, the importance of extracting high spatiotemporal resolution surface reflectance by fusing multi-satellite imagery has been growing. However, differences in observation geometry, spectral channels, and spatial resolution among satellites pose several technical challenges in the data fusion process. To overcome these limitations, multi-satellite imagery was preprocessed and corrected, followed by the application of image fusion and gap-filling techniques to generate high spatiotemporal resolution fused surface reflectance. The performance of this technique can be further enhanced by incorporating deep learning-based mechanisms, and the accuracy of the generated results can be evaluated through comparison with ground observation data. The fused images are expected to be utilized in various fields, including land use change monitoring, ecosystem assessment, and climate change research.