With the rapid advancement of Earth observation technologies, the availability of high-resolution remote sensing (RS) data has expanded across a wide range of domains, including environmental monitoring, land use classification, disaster management, and defense. However, the processing and interpretation of such data are increasingly challenged by large volumes, complex spatial-spectral structures, and limited labeled samples. Computational Intelligence (CI), inspired by natural and biological systems, offers promising strategies to address these complexities through adaptive learning, feature extraction, and decision-making.
This reprint explores the latest CI-driven methodologies applied to remote sensing tasks, as reflected in recent advancements such as hybrid retrieval systems, lightweight segmentation networks, few-shot classification models, semantic-enhanced image captioning, and dual-domain transformers for change detection. It highlights contributions from contemporary studies that tackle practical challenges in RS, including pan-sharpening, road extraction, radar imaging, and hyperspectral unmixing. By integrating theory with state-of-the-art applications, the reprint serves as a valuable reference for researchers and professionals working at the intersection of artificial intelligence and Earth observation.