Exjobb – Path planning and 3D mapping using Autonomous UAV with a RGB-D camera

Abstract

to perform automatic path planning and re-planning in initially unknown environment. Depth data from the RGB-D camera is used for position estimation of the vehicle as well as for mapping up the environment. Position estimation of the vehicle is done by using a rotation fused Iterative Closest Point (ICP) algorithm[P.J. Besl, 1992] and Box Approximation is used to map up a discrete obstacle map of the initially unknown environment. Path planning is performed on the flight using the obstacle map and the re-planning D* Lite algorithm[S. koening, 2002].

In this thesis a RGB-D camera was mounted on an unmanned vehicle with the goal We compare the use of Box Approximation and Signed Distance Function (SDF) for discretization purposes. Evaluates Point-to -Point ICP, Point-to-Plane ICP aw well as Translation Approximation for tracking purposes.

Thesis and article

Thesis
Article