Autonomous navigation system

• Main Author: Lai, Ming Hui
• Other Authors: Lam Siew Kei
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision
• Online Access: https://hdl.handle.net/10356/163034
• Institution: Nanyang Technological University

Visual SLAM has gained traction in the field of autonomous navigation systems in recent years. In the past, LiDAR SLAM reigned as the most popular SLAM metho in autonomous navigation, due to its highly accurate and fast performance. However, the expensive price of LiDAR causes a demand for lower cost but effective LiDAR solutions. Visual SLAM relies mainly on cameras, which are cheaper than LiDAR and have therefore gained popularity as a cost-effective alternative to LiDAR SLAM. There have since been many visual SLAM methods proposed by the community which have been demonstrated to be capable of accurate and robust SLAM, such as ORB-SLAM 3. In this project, we aim to create an autonomous navigation system using ORB-SLAM 3. We leveraged pre-existing robotic solutions provided in the Robot Operating System (ROS) framework and combined it with ORB-SLAM 3 to develop a navigation stack capable of autonomously navigating a robot in an indoor environment.