Low-Cost Autonomous Contour Localization and Area Measurement for Irregular Flood Zones Using a Small UAV

Timely and accurately locating the contours and measuring the areas for flood zones at low cost is of great significance for disaster assessment, relief and rebuilding. To compensate for the drawbacks of traditional localization and measurement modes for irregular flood zones, a novel low-cost solution for underdeveloped countries or areas using a small unmanned aerial vehicle (SUAV), which is independent of complex image processing approaches, is proposed. In this solution, first, the virtual obstacles (VOs) are generated based on the real detected ground images via a low-computational-cost image recognition algorithm. Then, based on the VO distribution, the SUAV plans the path to track and encircle the irregular contour so that the contour localization and area measurement can be realized. Finally, the proposed solution is verified by an experiment that introduces a real lake as the irregular flood zone and an SUAV model in reality to simulate the real flight situation. The results show that the localization and measurement errors are only 1.51% and 2.58%, which are higher than the manual measurement mode.