LiDAR Based Obstacle Detection for Unknown Environment

Hoh, Zhao Quan (2019) LiDAR Based Obstacle Detection for Unknown Environment. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text HohZhaoQuan.pdf Restricted to Registered users only Download (2MB)

Abstract

LiDAR technology is going to play a key role in the future for autonomous vehicles. There is already advancement for LiDAR based autonomous vehicle in urban environment, however LiDAR based autonomous vehicle in rural, unknown environment are yet to be exploited. Research in autonomous vehicles has seen dramatic advances in recent years, due to the increases in available computing power and reduced cost in sensing and computing technologies, resulting in much more advance level of autonomous vehicles. This project talks about the use of LiDAR surveying technology in autonomous vehicle and how it is implemented to survey and scan unknown outdoor environments. The algorithm which is the Rao-Blackwellized Particle Filter method is introduced to play a part in process the laser scan data and to create a map. Trajectory Rollout Algorithm is also implemented for path planning and carry out obstacle detection and avoidance. For this project, the process procedure is divided into three parts, scanning, mapping and also navigating. The results obtained show promising result where the Neato LiDAR is able to carry out laser scanning indoor, outdoor and also in dark environment. The Neato Robot was also able to carry out autonomous navigation using the path planning and obstacle avoidance algorithm mentioned above. As a conclusion, the XV-15 LiDAR sensor is capable of scanning indoor and outdoor environment, Using Gmapping module, the Neato robot is able to use the laser scan data to carry out map generation. Other than that, the Neato robot is also able to carry out autonomous navigation and obstacle avoidance using the Trajectory Rollout Algorithm for path planning and cost map of obstacle detection.