基于遥感图像分析的极区海冰漂移研究  

Study on the Polar Sea Ice Drift Based on Remote Sensing Image Analysis

作　　者：孙鹤泉[1] 李春花[2] 张志刚[1] 

机构地区：[1]海军大连舰艇学院军事海洋系,辽宁大连116018 [2]国家海洋环境预报中心,国家海洋局海洋灾害预报技术研究重点实验室,北京100081

出　　处：《海洋技术》2015年第1期10-14,共5页Ocean Technology

基　　金：国家自然科学基金资助项目(41376188);海洋公益性行业科研专项资助项目(201205007);极地考察专项资助项目(CHINARE-03-01);国家海洋局极地对外合作项目资助(IC201211)

摘　　要：对极区的海冰运动进行观测具有重要的意义,卫星遥感技术是目前极区海冰监测的重要手段。星载多光谱仪器和合成孔径雷达(SAR)能够生产高分辨率、宽刈幅、连续的遥感图像,是开展海冰遥感监测的优选载荷。通过对极区遥感图像的分析,可以获取极区海冰漂移运动的资料信息。文中提供了可用于极区海冰漂移研究的遥感资料源,详细介绍了两种基于遥感图像的海冰运动分析方法:相关匹配法与光流算法,以及两种方法的对比;利用文中介绍的遥感资料和分析方法,在两极海区发现了明显的海冰中尺度运动特征。It is of great significance to observe the sea ice motion in wide polar areas. With the motion and destruction of sea ice, the conventional in-situ instruments are limited for polar sea ice observations. Satellite remote sensing has become an important way to monitor the sea ice at high latitudes. Among a few types of the satellite payloads, multi-spectral imaging spectrometer and synthetic aperture radar（SAR） are preferred for monitoring the sea ice dynamics with their high resolution, wide swath and continuous imaging function. The sea ice dynamics in the ice-covered areas can be monitored by analyzing remote sensing images. The open data sources of remote sensing images are provided in the paper, with two kinds of algorithms based on remote sensing images proposed and compared in a detailed manner： the correlation matching method and optical flow algorithm.Through applying the remote sensing data and methods proposed in the paper, sea ice drift with mesoscale dynamic features can be discovered clearly in polar sea areas.

关 键 词：海冰漂移 遥感图像 相关匹配法 光流算法 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]