A Robot Mapping Technique for Indoor Environments

This paper presents a robot mapping technique that generates occupancy grid maps of the environments traversed by a robot using both 2D and 3D sensor data. The result of the mapping technique is a hybrid map of the environment. The decision to use 3D sensor data instead of 2D sensor data is based on the camera feed. The proposed technique is implemented in Robot Operating System (ROS) framework. Simulation worlds with different floor plans were created to generate the hybrid maps of the worlds using a mobile robot as a Unified Robotics Description Format (URDF) file equipped with 2D LiDAR sensor and 3D Kinect sensor.