摘要
针对已有的围填海图斑提取方法精度不高、普适性不强、自动提取结果不理想等问题,该文提出了通过构建归一化差异水体指数(NDWI)进行围填海变化图斑自动提取的方法。以高分辨率QuickBird影像和HJ-1卫星影像为数据源,首先,根据研究区的用海类型进行5种易混淆地物的波谱特征分析;然后,根据水体与非水体的光谱特征差异,构建2009、2011年两个时相的NDWI指数;最后,将两时相NDWI指数影像进行空间相减,设置判断阈值,识别围填海变化图斑,并以目视提取结果作为依据验证其自动提取效果。对比分析结果表明:利用该文构建的两期NDWI指数可以将大部分围填海区域准确、自动地探测出来,可以将该方法纳入到沿海地区围填海变化监测的业务中。
In order to resolve the problems of low accuracy, narrow applicability and poor result for exsiting automatic extraction method of coastal reclamation area, this paper devoted to develop NDWI （Normalized Difference Water Index）. And the high resolution QuickBird and HJ-1 images were used as data source. Firstly, the spectral character of 5 ground objects which are easy to confuse in remote sensing images were analyed. Secondly, the NDWI of 2009 and 2011 were computed separately according to the difference of spectral character between water and non-water. Finally, the coastal reclamation areas was obtained by subtracting NDWI2011 from NDWI2009, which was evaluated and verified by visual extration results. Experiment results indicated that most coastal reclamation areas could be extracted using two temporal NDWI images, and this method could play a role in change detection for coastal reclamation.
出处
《测绘科学》
CSCD
北大核心
2015年第6期42-45,76,共5页
Science of Surveying and Mapping
基金
中央高校基本科研业务费专项资金项目(2013QJ055)
中国土地勘测规划院业务项目(201309106308)
