水深测量
多光谱图像
遥感
卫星图像
卫星
多光谱模式识别
均方误差
地质学
地理
环境科学
海洋学
数学
统计
工程类
航空航天工程
作者
Rania Hassan,Ahmed Yousuf Saber,Sameh B. El‐Kafrawy,Mostafa Rabah
出处
期刊:Springer proceedings in earth and environmental sciences
日期:2023-01-01
卷期号:: 345-357
标识
DOI:10.1007/978-3-031-40447-4_39
摘要
Satellite imaging provides a whole spectral characterization of a region in digital type. These data are being employed for various vital applications at shallow coastal areas like Bathymetric information which is important for the applications of hydrological engineering like the processes of sedimentary and the studies of coastal. Multispectral satellite imagery has given a great coverage, low price and time-effective resolution for bathymetric measurements. The current study evaluates performance of 2 models to measure water depth within the south of Marsa Alam center—Red Sea Governorate on Halaib and Shalatin road. The models are neural network fitting algorithms (NN) and the bagged regression trees (BAG). Landsat 8 satellite imagery data was utilized to survey the execution of models. The used models were utilized to get the calculation of bathymetric maps in shallow coastal areas from multispectral satellite images using the reflectance values of (red, green) bands and the ratios of (green/red), and (blue/red) bands. The (BAG) resulted in RMSE 0.6219 m and R2 of 0.59 where (NN) yielded RMSE of 0.6911 m and R2 of 0.59 over shallow water depths. The BAG algorithm, produced the foremost reliable results by RMSE 0.6219 m and R2 of 0.59, tested to be the desirable algorithm for bathymetry calculation for study area.
科研通智能强力驱动
Strongly Powered by AbleSci AI