摘要
传统的海岸线调绘方法比较费时、费力,而且资金消耗巨大,因而会逐渐被遥感技术所取代。本文利用Landsat 8的OLI数据提取长江入海口处的九段沙水边线。由于水体、土壤等地物的光谱特征不同,所以可以选择不同的方法进行提取。本文选用了灰度阈值法、边缘检测法中的Soble算子和Roberts算子、改进的归一化水体差异指数(MNDWI)等方法对水边线进行提取研究,并对结果进行比较分析。发现Landsat 8中的第一波段有利于水边界线提取,边缘检测中的Soble算子提取的九段沙水边线可以得到较好的结果。
Traditional methods of coastline drawing costed more time, laboriousand huge capital consumption, which will gradually be replaced by remote sensing technology. In this paper, we use Landsat eight OLI data to extract nine sand waterlines at the edges of Yangtze River. Due to the different spectral characteristics of water, soil and other surface features, so we can choose different meth- ods to extract them. This paper selected the gray threshold, Soble operator and Roberts operator of the edge detection method, im- proved Normalized Difference Water Index （MNDWI） and other methods to carry out the research of waterline extraction, and the re- suits were compared and analyzed. We found that the first wave band of Landsat eight in favor of the water boundaries extraction, and use the Soble operator of edge detectionto extract nine sand waterlines can get better results.
出处
《测绘与空间地理信息》
2016年第10期28-32,共5页
Geomatics & Spatial Information Technology
