Automatic Construction of Envorinment Model for Simulator Gazebo from Sensor Data

Abstract

This paper describes a method for creating a 3D model for Gazebo, a robot simulator. In order to do a simulation in Gazebo, a user has to provide a 3D model of the environment in which robot is simulated. As the environment gets larger, the burden to prepare the 3D model increases. This research aims an automatic construction of 3D models from the data obtained by travel data of a mobile robot. The input point cloud is obtained using 3D LiDAR on a mobile robot and are converted into 3D mesh through the proposed mesh reconstruction process. The model is imported into Gazebo, and a simple experiment was performed to check the availability of the mesh for simulation, and the model was evaluated by comparing the pointcloud obtained in real world against the one obtained from simulation.