Small object detection in optical remote sensing images: current status and trends

Small object detection in optical remote sensing images holds significant research value and broad application prospects, spanning diverse fields such as military reconnaissance, disaster monitoring, urban management, and environmental protection. However, it confronts severe challenges due to factors like the complex backgrounds of remote sensing images and the difficulty in characterizing the features of small-sized objects. Currently, algorithms for small object detection in optical remote sensing images often adopt advanced general object detection algorithms as their base models, with optimizations made to the processing workflows and structural designs for input images. To address this, this paper first reviews the feature extraction methods commonly used in general object detection algorithms. Based on this foundation, it systematically organizes the intelligent detection algorithms for small objects. Finally, it discusses the key challenges in existing research and future development trends. This review aims to provide valuable references for researchers in related fields and promote continuous progress in the discipline.