Bathymetric estimation from satellite images using BEMD and FFT / 應用二維經驗模態分解及快速傅立葉轉換 於衛星影像推算海域地形之研究

碩士 / 國立交通大學 / 土木工程系所 / 102 / Bathymetric estimation was traditionally obtained using single/multi-beam echo sounder on boats. However, taking time-consuming, requiring high cost and precise surveying instruments are disadvantages of bathymetric surveying. In recent years, satellite images that provide multi-information of large area have been commonly used in a wide range of researches, including the investigation on wave-length and wave directions.
Those sub-images from one original satellite image are decomposed by bi-dimensional empirical mode decomposition (BEMD) and fast Fourier transform (FFT), to get spatial spectra where peak indicates the wave number and direction of primary waves. The track of wave directions from a sequence of sub-images forms a wave ray. The local water depth can be derived using dispersion relation of the linear wave theory from estimated wave-length and period of a sub-image.
The Fangliao coast of Pingtung county and the Toucheng coast of northern Ilan county are chosen for examining the accuracy of the proposed method by the root mean squared error between the measured and calculated water depths. Eight selected beach profiles are chosen to have a mean RMSE of 0.84m for the Fangliao coast using corrected wavelengths and a mean RMSE of 1.23m for the Toucheng coast using originally obtained wavelengths. It is proven that 2-m spatial resolution of chosen satellite images can produce an error of about 1 m in depth estimation using the proposed method. When the spatial resolution of satellite images is promoted someday, the proposed method can offer concise bathymetry and is applicable for coastal planning.

Identiferoai:union.ndltd.org:TW/102NCTU5015058
Date January 2014
CreatorsCheng, Yi-Lin, 鄭義霖
ContributorsChang, Hsien-Kuo, 張憲國
Source SetsNational Digital Library of Theses and Dissertations in Taiwan
Languagezh-TW
Detected LanguageEnglish
Type學位論文 ; thesis
Format60

Page generated in 0.0119 seconds