A Survey of Small Sea-Surface Target Detection for Maritime Search and Rescue

The detection of small surface targets plays a critical role in maritime search and rescue (SAR) operations, ensuring the safety of people and property at sea. This paper provides a comprehensive review of the latest advancements and research in small sea surface target detection for maritime SAR missions. Deep learning-based models facilitate accurate target detection and localization by transforming image or video frames into high-dimensional abstract representations, enabling effective detection in complex sea surface environments. However, challenges such as occlusion, blurring, and reflections on the sea surface significantly complicate small target detection. To address these challenges, this paper summarizes a range of effective approaches, including context information, multi-scale learning, anchor-free detection, super-resolution, attention mechanisms, and sample-oriented approaches. These approaches aim to enhance the performance of small target detection in applications such as uncrewed aerial vehicles (UAV) and uncrewed supply vessels. Furthermore, this paper classifies small target datasets, providing a detailed overview based on their collection methods and application scenarios, while highlighting representative datasets. Through a thorough analysis of both methodologies and datasets, this paper offers valuable insights and directions for the future development of small target detection technology in maritime search and rescue operations.