AUTOMATED OBJECT ANNOTATION USING FUSED CAMERA/LiDAR DATA POINTS

摘要

The present disclosure relates to computer systems and techniques for automatically annotating objects in map data used to drive autonomous vehicles. In general, the present computer system obtains LiDAR data points for the environment around the autonomous vehicle, projects the LiDAR data points onto image data, detects a target object in the image data, and a LiDAR corresponding to the detected target object. extract a subset of the data points, and register the detected target object in the map data if the extracted subset of the LiDAR data points meet the registration criteria, and drive the autonomous vehicle in the environment according to the map data .