Research on Obstacle Avoidance Strategies for Intelligent Connected Vehicles Based on Multi-Sensor Fusion and YOLOv5 Algorithm

This paper proposes a real-time obstacle avoidance strategy based on multi-sensor fusion with YOLOv5 algorithm. First, multi-sensor fusion is to combine different types of sensor data and process them together. Secondly, YOLOv5 algorithm is usually used for object detection, which optimizes the detection performance by unifying the processing of different data types. Currently, how to combine multi-sensor fusion and YOLOv5 to study the obstacle avoidance strategy of intelligent connected vehicles needs to clarify several issues: what is a reasonable multisensor fusion scheme; what are the advantages of the YOLOv5 algorithm in combining the data of each part; and how do the two work together to optimize the obstacle avoidance performance.