Abstract
Lake Water storage can be estimated based on Lake Boundary and lakebed terrain. The boundary of a lake can be extracted from the remote sensing satellite image or the Unmanned Aerial Vehicle (UAV) orthophoto, but the information of the lake and its surroundings, which is supposed to be three-dimensional, is recorded in two dimensions in this way. When the tall and dense vegetation are projected on the lake, the extracted lake boundary from orthophoto would be smaller than reality, which results in lower value of the lake water storage. UAV oblique photogrammetry can solve this problem by recording multi-angle images of the lake and its surroundings. Based on aerial photogrammetry theory, a real 3D model of the lake can be constructed by combing UAV oblique photogrammetric images. The model can accurately present the actual scene of the lake, and the extracted lake boundary can be very close to reality. In this study, UAV and Unmanned Surface Boats (USB) have been employed to collect UAV oblique photogrammetric images and lakebed terrain data, respectively. The lake boundary has been extracted based on the UAV real 3D model. The Digital Elevation Model (DEM) of lakebed terrain has been built based on lakebed terrain data collected by USB. Finally, the water storage of the lake is estimated based on the boundary and DEM. Experiments show that a real 3D model based on UAV data can depict lake boundaries accurately, which could eliminate the impact of vegetation shied around the lake. Utilizing the UAV and USB can be a standard method for estimating the water storages of small lakes.

This publication has 20 references indexed in Scilit: