基于SAR图像纹理的北极海冰厚度的反演研究  

ARCTIC SEA ICE THICKNESS RETRIEVAL BASED ON SAR IMAGE TEXTURE FEATURE

作　　者：于淼 卢鹏[1] 李志军[1] 石立坚[2,3] Yu Miao;Lu Peng;Li Zhijun;Shi Lijian(State Key Laboratory of Coastal and Offshore Engineering,Dalian University of Technology,Dalian 116024,China;National Satellite Ocean Application Service,Beijing 100081,China;Key Laboratory of Space Ocean Remote Sensing and Application,Beijing 100081,China)

机构地区：[1]大连理工大学海岸和近海工程国家重点实验室,辽宁大连116024 [2]国家卫星海洋应用中心,北京100081 [3]国家海洋局空间海洋遥感与应用研究重点实验室,北京100081

出　　处：《极地研究》2018年第3期329-337,共9页Chinese Journal of Polar Research

基　　金：国家重点研发计划专项(2016YFC1402702);国家自然科学基金面上项目(41676187;41376186);国家国际科技合作专项(2011DFA22260)资助

摘　　要：基于七景北极Radarsat-2 SAR图像以及中国第六次北极科学考察走航期间利用船侧录像观测获得的平整冰厚度数据,通过灰度共生矩阵计算纹理,确定了最适合反演海冰厚度的纹理参数。并分析了海冰厚度与纹理之间的相关关系,探讨了纹理反演海冰厚度的可能性。选取了最合适的纹理特征进行拟合,并利用所得经验方程进行反演验证,结果与实测数据吻合较好,平均相对误差13.7%。与传统的仅依靠后向散射系数反演海冰厚度进行对比,新方法的误差更小,证明了纹理特征反演冰厚的优势。A novel method to estimate ice thickness based on seven Arctic Radarsat-2 SAR images and data for level ice thickness collected from the 6th Chinese National Arctic Research Expedition(2014)is presented.Suitable values of texture parameters are confirmed using a gray level co-occurrence matrix(GLCM).Relationships between sea-ice thickness and texture features are analyzed,and the possibility of determining sea-ice thickness based on ice texture features is discussed.Curve-fitting equations can be determined by optimum texture features.Estimated thickness agrees well with in-situ measurements,with an average relative error of 13.7%.Compared with another commonly used method that depends on a backscattering coefficient,this new method has less error,indicating texture features can be reliably used to determine ice thickness.

关 键 词：北极 海冰厚度 灰度共生矩阵 纹理 

分 类 号：P731.15[天文地球—海洋科学] P714