Point cloud data processing method and apparatus, vehicle, and storage medium

Abstract

The present disclosure is a point cloud data processing method, device, vehicle and storage medium, which relates to the application of data processing technology in the field of vehicles. The method comprises: obtaining point cloud data to be processed containing semantic labels; based on the preset semantic labels and the preset first reflection intensity threshold, filtering and generating a first region from the point cloud data to be processed; based on the geometric features of the first region, dynamically expanding the first region and filtering to obtain a second region; based on the preset second reflection intensity threshold and the geometric constraint condition, filtering out the inflation noise points from the second region to obtain the corrected point cloud data. The present disclosure can effectively and automatically detect inflation noise points, avoid errors caused by misjudgment of the subsequent vehicle perception module, and improve the safety and reliability of the intelligent driving system.

Images