Nearshore bathymetry estimation using UAV

Abstract

An unmanned aerial vehicle (UAV)-based system is developed to flexibly survey nearshore bathymetry with high resolutions in both time and space. The system is based on the inverse estimation of water depth from the speed of wave crest lines extracted from video images captured by a UAV hovered at a height of 150 m. The extraction of wave crest lines was achieved by cross-shore distribution of the brightness for unidirectional waves while two-dimensional wavelet transform was applied for multidirectional waves. Only the data for long-period waves were used to ensure the accuracy of bathymetry estimation. The validity of the system was confirmed at three beaches with different characteristics: Yotsugoya beach with multiple sand bars, Ariake beach with submerged structures, and Asaba beach with sand bypassing.