基于多光谱数据的南海浅水水深测量算法研究  

Study of Shallow Water Depth Measurement Based on Multi-Spectral Data in South China Sea

作　　者：朱金山[1,2] 宋珍珍 纪轩禹 赵露露 ZHU Jinshan;SONG Zhenzhen;JI Xuanyu;ZHAO Lulu(College of Geomatics,Shandong University of Science and Technology,Qingdao 266590,China;Key Laboratory of Surveying and Mapping Technology on Island and Reef,National Administration of Surveying,Mapping and Geoinformation of China,Qingdao 266590,China)

机构地区：[1]山东科技大学测绘科学与工程学院,山东青岛266590 [2]海岛(礁)测绘技术国家测绘地理信息局重点实验室,山东青岛266590

出　　处：《测绘与空间地理信息》2019年第1期10-13,17,共5页Geomatics & Spatial Information Technology

基　　金：国家自然科学基金(41771408);山东省自然科学基金(ZR201702210379);2015测绘地理信息公益性行业科研专项(201512034)资助

摘　　要：沿海地区的水深测量是大多数海岸工程和沿海科学应用的常见要求。然而,在我国南海浅水区域(0—10 m)及部分政治敏感区,船只很难到达导致测量非常不便,这种情况下遥感测量方法凸显了其价值。本文基于QuickBird多光谱遥感影像及同时期多波束实地测量水深点数据,利用6种算法模型定量反演了中国南海甘泉岛周边浅海的水深。精度验证结果表明,基于遗传算法优化的BP神经网络模型与实际测量水深的相关系数R2大于0.96,测量结果可靠。Water depth measurement in coastal areas is a common requirement for most coastal engineering and coastal scientific applications.However,it is very inconvenient for ships to measures in the shallow waters of South China Sea(0-10 m)and some of the politically sensitive areas,.In this case,the remote sensing measurement method highlights its value.This paper based on QuickBird multispectral remote sensing images and simultaneous multi-beam spot measurements,and the water depth of shallow sea near Ganquan Island in South China Sea was quantitatively determined by six algorithms.The accuracy verification results show that the correlation coefficient R^2 between the BP neural network model based on genetic algorithm optimization and the actual measured water depth is greater than 0.96.

关 键 词：QUICKBIRD 水深测量 多光谱 BP神经网络 遗传算法 

分 类 号：P234.2[天文地球—摄影测量与遥感]