摘要
埃默里冰架(Amery ice Shelf,AIS)是南极洲第三大冰架,冰架状态影响着南极洲物质平衡和海平面变化,但目前对于AIS与海水交界的冰架前端位置确定研究甚少。基于哨兵一号(Sentinel-1)合成孔径雷达(synthetic aperture radar,SAR)影像提出了一种高效且精确的冰架前端自动检测方法,利用冰架和海水之间的过渡带的SAR后向散射系数分布特点,利用Sentinel-1SAR影像并结合单元最小恒虚警率(smallest ofconstant false alarm rate,SO-CFAR)和形态学滤波得到冰水二值图,采用滑动窗口和累积和的方法自动提取每条剖面线对应的冰架前端点位置,自动绘制AIS前端轮廓线。考虑SAR影像空间分辨率和剖面分辨率等因素对前端检测的影响,进行冰架前端参数优化,并分析有无浮冰对冰架前端提取精度的影响。为了验证影像空间分辨率对各种方法检测结果的影响,将AIS前端无碎冰的影像进行双线性内插法重采样处理,并与基于标准差与五大值法的冰架前端提取算法进行精度对比分析。实验证明提出的剖面法具有一定的适用性。此外,通过分析AIS前端有无碎冰发现,基于SO-CFAR和形态学滤波算法相结合的剖面法对冰架前端提取精度最佳,最优检测精度小于1个像素,且受表面融水、冰架破碎等较小,具有较强的场景适应性。
The Amery ice shelf(AIS) is the third largest in Antarctica. Its behavior controls the mass balance of Antarctica and affects sea level rise while studies focusing on the calving at the boundary of AIS are few. In this paper, an efficient and automatic algorithm for AIS frontal line detection using Sentinel-1 SAR imagery is proposed. Based on characteristics of SAR(synthetlc aperture radar) backscatters distribution, smallest of constant false alarm rate(SO-CFAR) algorithm and morphological filter is combined to obtain the binary map of the Sentinel-1 SAR imagery. Then, sliding windows and cumulative sum strategies are applied to automatically detect frontal positions and map the ice frontal lines. The sensitivity of spatial resolution and profile resolution is analyzed by bilinear interpolation for image resolution resampling, and affection of floating ice is considered in the experimental analysis section. Experimental results are compared with state-of-the-art algorithms, and accuracies reported that the proposed algorithm can provide an accurate frontal line under one pixel and prove it is suitable for ice surface melting water and calving conditions.
作者
李斐
王振领
张宇
张胜凯
朱婷婷
LI Fei;WANG Zhenling;ZHANG Yu;ZHANG Shengkai;ZHU Tingting(Chinese Antarctic Center of Surveying and Mappmg,Wuhan University,Wuhan 430079,China;State Key Laboratory of Information Engineering in Surveying,Mapping and Remote Sensing,Wuhan University,Wuhan 430079,China)
出处
《武汉大学学报（信息科学版）》
EI
CSCD
北大核心
2018年第12期2012-2022,共11页
Geomatics and Information Science of Wuhan University
基金
国家重点研发计划(2017YFA0603104)
国家自然科学基金(41801266,41531069)
中央高校基本科研业务费专项(2042018kf0050)
博士后面上项目(2017M612512)~~
