Low-cost reality capture system for high-fidelity unreal engine based robot simulator

• Main Author: Zheng, Qichun
• Other Authors: Xie Lihua
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Electrical and electronic engineering
• Online Access: https://hdl.handle.net/10356/167187
• Institution: Nanyang Technological University

Testing algorithms for robots and unmanned aircraft can be costly and even dangerous. Simulators can be used to provide a virtual environment for efficient and safe testing. Realistic reproductions of the real world are preferred when building the simulators, as they can help to minimize the efforts required to migrate from simulators to actual robots. Currently, the environments in simulators are mostly made by expensive manual modelling. This project aims to provide a high-fidelity reality capture system at a low cost that is essential for developing control algorithms. The proposed reality capture system should meet two key requirements: (1) reproduce a realistic environment, and (2) at a reasonable cost. This project utilizes a Light Detection and Ranging (LiDAR) sensor, a camera, and an Inertial Measurement Unit (IMU). The components were assembled to form a handheld scanner. The data collected by the sensors were processed by multiple software to produce the desired format as the virtual environment for simulators. Finally, a sample simulator was created to verify the entire process. The main contributions of this project are (1) the use of affordable sensors for data collection, (2) the process to convert raw data to meshes, and (3) the acceptable quality of the virtual environment.