首页 | 官方网站   微博 | 高级检索  
     

海面溢油无人机高光谱遥感检测与厚度估算方法
引用本文:任广波,过杰,马毅,罗旭东.海面溢油无人机高光谱遥感检测与厚度估算方法[J].海洋学报,2019,41(5):146-158.
作者姓名:任广波  过杰  马毅  罗旭东
作者单位:自然资源部第一海洋研究所,山东 青岛,266061;中国科学院烟台海岸带过程研究所,山东 烟台,264003;广州星博科仪有限公司,广东 广州,510070
基金项目:国家自然科学基金项目(61601133,61890964,41576032,41706208)。
摘    要:海上溢油是海洋国家所面临的共同问题,但至今仍没有一种可靠实用的海上溢油准确识别和油量遥感监测方法。为此,本文以无人机高光谱遥感为手段,开展了海面溢油检测与厚度估算方法研究。实验中,通过搭建室外大型水槽溢油实验装置,获取了模拟真实海洋环境条件下不同溢油量的遥感和现场光谱数据,在此基础上,分析并提取了海上溢油特征光谱波段,给出了海上溢油高光谱检测模型;针对现场实验条件下水面油膜厚度难以测定的问题,设计了3种利用总体溢油量的油膜厚度估算模型。得到如下主要结论:(1)675 nm和699 nm是海上溢油检测的有效特征波段,但对极薄的油膜没有检测能力;(2)提出了归一化溢油指数模型、反比例模型和吸收基线模型等3种海上溢油油膜厚度估算模型,其中对于薄油膜(厚度≤ 5 μm)和厚油膜(厚度>50 μm),反比例模型是溢油厚度反演的首选也是唯一选择。对于中厚度油膜,晴朗天气条件下,归一化溢油指数模型是油膜厚度反演的首选,同时反比例模型和溢油吸收基线模型也都有较好的反演能力,而在多云天气条件下,反比例模型效果最佳。

关 键 词:溢油检测  溢油厚度评估  无人机高光谱  溢油遥感
收稿时间:2018/5/6 0:00:00

Oil spill detection and slick thickness measurement via UAV hyperspectral imaging
Ren Guangbo,Guo Jie,Ma Yi and Luo Xudong.Oil spill detection and slick thickness measurement via UAV hyperspectral imaging[J].Acta Oceanologica Sinica (in Chinese),2019,41(5):146-158.
Authors:Ren Guangbo  Guo Jie  Ma Yi and Luo Xudong
Affiliation:1.First Institute of Oceanography, Ministry of Natural Resources, Qingdao 266061, China2.Yantai Institute of Coastal Zone Research, Chinese Academy of Sciences, Yantai 264003, China3.NBL Imaging System Ltd., Guangzhou 510070, China
Abstract:Oil spill is a common problem faced by marine countries, but there is still no reliable and practical method for oil slick accurate identification and quantity measurement via remote sensing technology. Based on the UAV hyperspectral imaging experiment, methods of oil spill detection and thickness estimation are studied. In the experiment, the UAV hyperspectral remote sensing and field spectral data of oil spill with different quantities are obtained in an oil spill experiment tank of large outdoor flume under the condition of simulating real marine environment. Then the feature spectral bands based oil spill detection and oil slick thickness estimation models are found. At last we get the following conclusions:(1) 675 nm and 699 nm are the effective characteristic bands of oil spill detection, however, they have no detection capability for the very thin oil slick (thickness ≤ 5 μm), (2) 3 kinds of oil slick thickness estimation models witch are Normalized Difference Oil Spill Index (NDOSI) model, inverse proportion model and absorption line height model are proposed, in which the inverse ratio model is the only choice for thin and thick(thickness>50 μm) oil slick. For the medium thickness oil slick, the NDOSI model is the best choice, and the inverse scale model and the oil spill absorption baseline height model have better inversion ability, and in cloudy weather, the inverse scale model is the best.
Keywords:oil spill detection  oil slick thickness estimation  UAV hyperspectral  oil spill remote sensing
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《海洋学报》浏览原始摘要信息
点击此处可从《海洋学报》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号