mepix / RoboND-T2P3-RTABmapping

Udacity RoboND RTABmapping

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

RoboND-T2P3-RTABmapping

Overview

Mapping is an important part of robot navigation, whether maps are preloaded onto a platform for Monte Carlo Localization or as part of a SLAM pipeline. This project applies Real Time Appearance Based (RTAB) Mapping to map two different environments, an interior Kitchen World and an exterior Custom Firehouse Environment. Both worlds could be successfully mapped in simulation using a differential drive robot with an RGBD camera and a laser scanner. For each world, 2D and 3D maps of the environment were created. Loop Closure was achieved 303 times for the Kitchen World and 27 times for the Custom Firehouse Environment.

UDACITY: Kitchen World Environment

2D Map

Kitchen 2D

3D Map

Kitchen 3D

Database Viewer with Loop Closures (Purple)

Database Viewer

CUSTOM: Fire House World Environment

2D Map

Firehouse 2D

3D Map

Firehouse 3D

About

Udacity RoboND RTABmapping


Languages

Language:CMake 56.6%Language:Python 43.4%