Research on Multi-sensor Fusion Traffic Environment Perception Method for Autonomous Driving

In the field of autonomous driving, image generation of multimodal perception data plays a key role in understanding the data, validating algorithms, analyzing system behavior, and effectively communicating with others. Based on ROS (robot operating system) and computer vision algorithm evaluation dataset (KITTI dataset) in autonomous driving scenarios, this study proposes a comprehensive method for generating multimodal perception data, trajectory and workshop distance images using image data, point cloud data, IMU (inertial measurement unit)/GPS (Global Positioning System) and other data. Finally, this diverse and complex data is visualized accurately and efficiently to visualize the vehicle’s surroundings and behavior.