Visual asset inspection using precision UAV techniques

Recent years have seen a surge of interest in the use of UAVs for asset surveying applications including chimney stacks, power lines, wind turbines and other structural assets. Most commercial systems available are manual, relying on a human operator to fly the UAV or drone, and a manual process of image stitching and analysis to check for faults and defects. Errors in the position of the inspection drone make guaranteeing full coverage and accurate defect registration particularly problematic with this approach. This paper details our work at University of Strathclyde, conducted under the UK Research Centre in Non-Destructive Evaluation, in areas that are essential to improving the accuracy of such drone based structural asset inspection; we describe this set of approaches for improving the ability of drones to make such measurements. We present the results of a study aimed at better understanding of the positional accuracy and stability of UAV’s when used in indoor and outdoor environments for inspection purposes, utilising high precision laser tracking and photogrammetric approaches. Coverage path planning approaches were implemented to ensure full area coverage and provide accurate registration of faults. Finally, using image stitching – it is possible to get both full area visualisation of the structure, and more importantly to reconstruct a 3D CAD model of the surface of the structure under inspectio