Development of an intelligent mobile robot localization system using Kinect RGB-D mapping and neural network

The most important issue of intelligent mobile robot development is to navigate autonomously in the environment for completing certain task demands. Nowadays, the Kinect sensor is affordable and popular for acquiring environment RGB image pixels with depth estimation. In this study, we focus on developing the indoor localization system for intelligent mobile robot applications. The innovation of this research is to combine the RGB-D mapping and neural network training for achieving an Indoor Positioning System. It is expected that the inputs are the robot's observations of environmental features / landmarks and the direct output is the robot's posture which will correspond to the RGB-D map. All the experimental results suggest that the robot's posture and localization adjusts very efficiently with this study's proposed method.