On-Orbit Fast Small Target Detection Method Based on Improved YOLOv5 Model

Small target detection in remote sensing images is a challenging research field. Real-time on-orbit small target detection is of great significance for rapid discovery and tracking of remote sensing targets. In this paper, the feasibility of automotive-grade system on chips (SOC) to achieve on-orbit high-performance computing is analyzed, a high-performance computing solution that meets the complex space application environment is proposed. At the same time, based on the improved YOLOv5s model and the block parallel processing framework, a real-time on-orbit small target detection method for remote sensing images is proposed. Experimental results show that the small target detection time of this method is about 100ms for a 5k×3k resolution image, meanwhile, the actual results for small object detection are also better than the original YOLOv5 model.