Road Guideline Detection Method Based on E-YOLOv5 Algorithm in Autonomous Driving

To address the performance limitations of traditional road surface guideline detection models, this study proposes an optimised detection model based on the YOLOv5 algorithm. By integrating the convolutional block attention module and optimising anchor box parameters, the model enhances feature extraction capabilities for small targets, thereby improving detection accuracy and real-time performance. Experimental results demonstrate that the improved model exhibits superior performance in both simulated and real-world scenarios, effectively mitigating issues such as missed detections, false positives and positional deviations. It significantly enhances robustness in complex environments. This research provides an efficient and reliable vision-based detection solution for autonomous driving and smart logistics, advancing the practical application of computer vision technologies.