Three-Dimensional Trajectory Design for Post-disaster UAV Video Inspection

Abstract

This paper considers unmanned aerial vehicle (UAV) trajectory design for post-disaster information collection based on video inspection. Unlike the conventional UAV photogrammetric applications that aims to cover the whole area of interest, we model the critical locations in the disaster area as points of interest (PoIs) with heterogeneous resolution requirements. An efficient algorithm is then proposed to design the three-dimensional (3D) UAV trajectory that minimizes the time required for inspecting all the PoIs, subject to the image quality constraints, in terms of the image resolution, produced blur and the number of shots for each PoI. The proposed modeling and trajectory design can greatly reduce the time required for UAV video inspection, which hence enables fast response and rescue planning after the disaster.