Small Target Traffic Sign Detection Method Based on YOLOv8s

To tackle the challenge of low detection accuracy for traffic signs in complex road conditions, we propose a YOLOv8s based detection method. This method improves the original model by introducing a novel network architecture that enhances small object detection. A three-branch feature extraction module is also designed to merge edge, local, and global information, boosting the model’s performance in complex environments. Experimental results on the augmented TT100k + dataset show that the improved algorithm increases precision, recall, and mean average precision by 1.0%, 6.9%, and 4.9%, respectively, compared to the original algorithm. These improvements provide strong support for autonomous driving and intelligent transportation systems.