Enhancing Autonomous Vehicle Navigation at Complex Junctions Using LIDAR and YOLOv5-Based Detection

Vehicle autonomy demand continues to grow because AI developments and sensor technology enable vehicles to navigate schematically while protecting safety conditions. The main obstacle for autonomous vehicles involves handling advanced junctions since their decision-making system needs to remain agile and precise. The study implements a sensing system including LIDAR depth sensors and camera-based traffic signal detections while YOLOv5 is employed as its deep learning identification system. The method employed transfer learning strategies to reduce training duration without compromising detection quality. We use a component integration system to maintain reliable performance across different conditions e.g. low visibility situations. The proposed autonomous vehicle infrastructure has been tested in both simulated and real-world environments. The system demonstrates superior performance in junction and obstacle avoidance operations over traditional autonomous vehicle technologies.