红外超光谱谱指纹检测和识别海洋溢油污染方法研究

项目名称： 红外超光谱谱指纹检测和识别海洋溢油污染方法研究

项目编号： No.61275098

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 无线电电子学、电信技术

项目作者： 马泳

作者单位： 华中科技大学

项目金额： 75万元

中文摘要： 本项目针对遥感监测海洋溢油污染中，无法实时确认是否为溢油污染这一问题，给出一种基于红外全谱段超光谱遥感探测和识别海面溢油污染的新方法。论证该方法不仅具有全天时、快速实时探测能力，而且具有识别确认溢油污染的能力。研究从海水的固有光学特性以及海面溢油表征光学特性的理论研究入手，研究和分析海面溢油的红外反射和辐射光谱特性这一科学问题。其次研究遥感光谱曲线处理中的若干关键技术，如大气校正技术、噪声估计和去除技术、光谱特征提取技术等，实现对遥感光谱曲线的精确还原。通过理论研究结合实际试验获取的红外超光谱数据，在获得信源模型和数据的基础上，建立典型溢油污染物的特征数据库。最后论证采用红外全波段高分辨率傅里叶变换光谱仪，通过遥感方式获得海面溢油的红外反射光谱和辐射光谱的可行性和精确性。

中文关键词： 红外遥感；海洋监测；红外超光谱；溢油污染；探测识别

英文摘要： Current methods for the remote sensing of ocean oil spill have their disadvantages as they fail to recognize the oil spill pollution in real time. In order to address this problem, a novel method for the remote sensing of oil spill pollution based on infrared hyperspectral technique is proposed in this project. The proposed method is able to recognize the oil spill pollution in real time both in day and at night, as well it provides the ability to identify the type of the spilled oil. The project includes the following research: first, based on optical and radiation theories of the ocean water and oil, the properties of infrared reflective and radiation spectrum are studied. Then critical signal processing technologies such as atmospheric correction, de-noising method and spectral feature extraction are also studied to realize the accurate retrieval of remote sensed spectrum. Furthermore a hyperspectral library of typical oil spills is built by theoretical study and experiment analysis. Finally, the precision of the obtained oil spill hyperspectral data is studied to evaluate the availability of the proposed method.

英文关键词： infrared remote sensing；ocean mointoring；infrared hyperspectral；oil spill pollution；detection and identification