Exploration and Mapping of Warehouses Using QuadrotorHelicopters

Abstract: The goal of this thesis work is mapping of warehouse infrastructure using AR.Drone, aquadro helicopter equipped with a wide angle camera as its main sensor for mapping.Parallel Tracking and Mapping algorithm is employed for localization and creationof 3D point cloud of corner feature in the environment. Our approach for mapping was adata fusion algorithm that combines point cloud data from PTAM with estimated pillarpositions extracted by multi-stage image analysis algorithm.The methods implemented in ROS and Matlab has been successfully tested in real warehouseenvironment. The system is capable of creating an accurate map if some factorsare controlled. Details of results and analysis of methods are provided.