Satellite Remote Sensing Data Fusion Technology: A Review

Satellite remote sensing data fusion has become a key technology in the field of geospatial sciences, enhancing accuracy and enriching information by integrating multisource datasets. This review aims to provide a comprehensive examination of state-of-the-art satellite data fusion techniques. From the perspective of satellite platforms and their onboard sensors, satellite remote sensing data fusion is categorized into four types: single-satellite multiband fusion, homogeneous multisatellite multisensor fusion, heterogeneous multi-satellite multisensor fusion, and nontraditional sensor fusion. Compared to the traditional classification of satellite remote sensing image fusion into pixel-level, feature-level, and knowledgelevel fusion from a machine vision perspective, this categorization offers a deeper insight into the characteristics of satellite remote sensing image fusion and the corresponding application scenarios for different fusion technologies. By providing an in-depth overview of current satellite remote sensing image fusion techniques and their development, this paper aims to guide researchers and practitioners, thereby promoting the rapid advancement of applications utilizing satellite remote sensing data.