Autonomous driving data annotation is an indispensable part of the research and development of autonomous driving technology. It involves the detailed annotation of a large amount of real-world data to train the perception, decision-making and execution capabilities of the autonomous driving system. Simply put, it is to The information that autonomous driving systems need to recognize and understand is extracted from the raw data and tagged in a specific way so that the system can learn and recognize it.