基于多特征的人工海岸线提取方法

Coastline extraction from remote sensing images based on multi-features approach

在中高分辨率的遥感影像中,海岸地物呈现出更多的细节,尤其是受人类影响程度较大的人工海岸。文中针对较为复杂的人工海岸遥感影像,进行基于多特征的海岸线提取方法研究。实验选用SPOT影像作为实验影像,分析实验区海岸带SPOT影像的光谱特征、形状特征和空间关系特征,使用多种特征来提取海岸线,实验结果表明,方法能对海岸带地物进行正确分类并实现海岸线的自动提取。

Coastal features show more details in medium-high resolution remote sensing images, especially the artificial coast which is influenced greatly by human beings. A remote sensing image of complex artificial coast is studied in multi-features ways. An experimental image of SPOT is analyzed by spectral, shape and spatial relations feature extracting. The experimental results show that this method can classify the features of the coastal zone correctly and can extract the coastline in an intelligent way.