Robot Mapping and SLAM: Map My World Robot
Udacity Robotics Software Engineer Nanodegree Program
Index Terms — Robot, Mapping, SLAM, GraphSLAM, RTAB-Map.
Abstract
A 2D occupancy grid and 3D octomap are created from a provided simulated environment. Furthermore, the own simulated environment to map is created as well. RTAB-Map, which is the best solution for SLAM to develop robots that can map environments in 3D, is used. The mobile robot model with a RGBD camera and a laser sensor is created.
INTRODUCTION
In Simultaneous Localization and Mapping (SLAM), a robot constructs a map of the environment, while simultaneously localizing itself relative to the map. The problem is challenging because neither the map nor the robots poses are provided.
RTAB-MAP (real time appearance based mappimg) is a GraphSLAM implementation. For this project, the rtabmapros package, which is a ROS wrapper (API) for interacting with rtabmap, is used. Using the created ROS package, both a 2D and 3D map of the supplied environment are created. From there then the own environment is built and a 2D and 3D map of that environment are generated. This project is completed on Jetson TX2.
