LiDAR Sensor Perception System for Self-Driving Cars

摘要

Present and future LiDAR sensor technology is the core role in the perception system for self-driving cars. Currently most of the automotive industries, academicians and researchers are using LiDAR sensor technology for perception systems. It is a remote sensing technology which creates 3D point cloud maps of surrounding environments and converts them in to 2D depth maps. LiDAR has a minimum of 16 channels for the application for object detection and navigation systems for self-driving cars. The objective of this research is to analyse the LiDAR sensor perception system in edge case scenarios at different weather conditions. For this research work 64 channels of Velodyne LiDAR punk sensor is used. It is integrated with the TurtleBot device with the help of a Robotic operating system - Noetic for controlling the device and Nvidia GPU GeForce GTX 770. User cases as the light and dark operating conditions are assessed and discussed.