Ship Target Detection in SAR Images Based on Multi-Scale Dynamic Fusion Model of YOLOv8

• DOI: 10.54097/r978ba16
• Publication Date: 2025-03-31T07:49:21Z
• Authors: Weihe Huang; Yinhong Song

Target detection is an important part of radar image interpretation. The use SAR technology combined with deep learning network models has become means ship target detection. Compared traditional methods, powerful data processing and feature extraction capabilities. Therefore, this paper proposes a multi-scale dynamic fusion (DFS) model for based on YOLOv8. includes three parts: fusion, alignment, attention mechanism enhancement. Finally, experiments were conducted the SAR-ship dataset. experimental results show that reach 96.4%.