基于多源遥感数据的冰间水道识别
首发时间:2017-09-26
摘要:采用MODIS可见光反射率、热红外亮温和Radarsat-2双极化后向散射等多源数据,通过建立决策树综合判断来识别波弗特海域冬季的冰间水道,并对识别方法进行评价。实验表明,MODIS热红外只能提取薄冰覆盖的线状冰间水道,而高分辨率的Radarsat-2影像可以提供更多冰间水道内的海冰类型信息,有助于识别多年冰、一年冰和重冻结的冰间水道等多种冰型。结合Radarsat-2后向散射和MODIS光学反射率等数据建立的决策树,能够利用多源数据各自的特点,综合判断海冰的类型从而提取冰间水道。决策树可以提供更准确的冰情信息,识别精度优于单变量方法。
关键词: 冰间水道 Radarsat-2 冰面温度 多源数据
For information in English, please click here
Sea ice leads detection based on multisensory remote sensing data
Abstract:Sea ice in Beaufort Sea is classified and leads extracted based on multisensory data, including optical and thermal images from MODIS and dual-polarization SAR image from Radarsat-2. Accuracy of the result is evaluated. Decision tree is used to incorporate multisensory data. Test result show that thermal images from MODIS is only useful in extracting lead covered by thin ice, while the high-resolution SAR image could provide more information on sea ice type in lead area, especially, refrozen ice lead. Advantages of different data were combined in decision tree to generate an optimal result of lead distribution, which could provide more information about the ice condition in leads. Overall accuracy of the result from decision tree is 14.8% higher than that from supervised classification of Radarsat-2 images.
Keywords: Sea Ice Lead Radarsat-2 Ice Surface Temperature Multisensory data
引用
No.****
动态公开评议
共计0人参与
勘误表
基于多源遥感数据的冰间水道识别
评论
全部评论0/1000