Visual inertial SLAM based autonomous navigation system

• Main Author: Chua, Eng Soon
• Other Authors: Lam Siew Kei
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision
• Online Access: https://hdl.handle.net/10356/171846
• Institution: Nanyang Technological University

Development and use of Visual and Visual Inertial SLAM for autonomous navigation have been prevalent for the past decade. Autonomous navigation in the past commonly used traditional SLAM with LiDAR as sensors which provides accurate depth ranging of the environment. The development of Visual SLAM using cameras as sensors was initiated as a lower-cost alternative to the usage of more expensive LiDAR for autonomous navigation. Today newer Visual SLAM methods with the use of Inertial Measurement Units (IMU) referred to as Visual Inertial SLAM and compatibility with different camera types such as stereo cameras and RGB-D cameras were proposed and shared with the public. In this project, we aim to implement an autonomous navigation system using COVINS and Multi-Robot Coordination. The navigation stack of Multi-Robot Coordination is integrated with COVINS, and the integrated system is then implemented on an Unguided Ground Vehicle to perform autonomous navigation in an indoor environment.