Route planning for angle constrained terrain mapping using an unmanned aerial vehicle

Unmanned Aerial Vehicles (UAVs) equipped with downward-facing, low-cost cameras can be used for terrain mapping. Using a photogrammetric technique, structure from motion, one can create accurate models for less cost than current methods. During mapping, images at certain locations (points of interest (POIs) need to be taken from specified directions due to environmental conditions. As the area that needs to be mapped increases the number of POIs increase. Therefore, we need a route planning algorithm that can determine efficient paths taking kinematic constraints of the vehicle and the angle of arrival at the POIs into account. To determine an efficient solution, we cast the routing problem as a Traveling Salesman Problem (TSP) with angular constraints, and develop two new solutions based on multi-lookahead and multi-neighbor strategies. Simulations are carried out to evaluate the performance of the algorithms.