Rectification of bowl-shape deformation of tidal flat DEM derived from UAV imaging

Abstract

It is necessary to periodically obtain topographic maps of the geographical and environmental characteristics of tidal flats to systemically manage and monitor them. Accurate digital elevation models (DEMs) of the tidal flats are produced while using ground control points (GCPs); however, it is both complicated and difficult to conduct GPS surveys and readings of image coordinates that correspond to these because tidal flat areas are not easy to access. The position and distribution of GCPs affect DEMs, because the entire working area cannot be covered during a survey. In this study, a least-squares height-difference (LHD) DEM matching method with a polynomial model is proposed to increase the number of DEM grids while using a presecured precise DEM to rectify the distortion and bowl effect produced by unmanned aerial vehicle (UAV) images. The most appropriate result was obtained when the translation parameters were quadratic curve polynomials with an increasing number of grids and the rotation parameters were constant. The experimental results indicated that the proposed method reduced the distortion and eliminated the error caused by the bowl effect while only using a reference DEM.