Traditional autonomous vehicles rely on a single sensor for environmental perception, which has issues such as insufficient information and susceptibility to interference, and cannot meet the requirements of complex road scenarios. Therefore, a Multi-Sensor Fusion (MSF) obstacle avoidance method has been proposed, which integrates information from various sensors including LiDAR, cameras, and millimeter-wave radar, utilizes high-precision maps and real-time obstacle information, and combines fuzzy reasoning mechanisms and adaptive fusion algorithms. Actual road tests have shown that the difference between the sensors based on the fusion algorithm and the true results is between 0.07 m and 0.23 m, improving the measurement accuracy by 0.61% and 0.41% compared to single cameras and single LiDAR. Multi-sensor fusion technology can maintain a safe stopping distance of ± 0.7 m, enhancing safety in emergency situations. In lane-changing tasks, the trajectory of the fused data is closer to the center of the lane, advancing by 1.3 m, increasing safety margins, efficiency, and accuracy. Compared with other fusion methods, the detection accuracy of the proposed method was improved by 51.52% and 30.43%, respectively. The results indicate that the multi-sensor fusion perception method for vehicle environment proposed in this study can improve the environmental perception capability of autonomous vehicles, bringing new opportunities and challenges for the development of future intelligent transportation systems, and has significant research significance and application value.
Multi-sensor Fusion Perception of Vehicle Environment and its Application in Obstacle Avoidance of Autonomous Vehicle
Int. J. ITS Res.
International Journal of Intelligent Transportation Systems Research ; 23 , 1 ; 450-463
2025-04-01
14 pages
Article (Journal)
Electronic Resource
English
Multi-sensor fusion , Perception , Vehicle environment , Autonomous driving , Vehicle obstacle avoidance Information and Computing Sciences , Artificial Intelligence and Image Processing , Information Systems , Engineering , Electrical Engineering , Automotive Engineering , Robotics and Automation , Computer Imaging, Vision, Pattern Recognition and Graphics , Civil Engineering , User Interfaces and Human Computer Interaction
Obstacle avoidance perception processing for the autonomous land vehicle
Tema Archive | 1988
|Predictive Obstacle-avoidance Control for Autonomous Vehicle
British Library Online Contents | 2015
|Multi-sensor fusion target vehicle automatic driving obstacle avoidance method and system
European Patent Office | 2021
|European Patent Office | 2022
|