基于天绘一号卫星多光谱影像的海岸线自动提取方法

A Method of Automatic Coastline Extraction Based on Multi-spectral Image of TH-1

针对常规人工海岸线测量方法耗费人力物力、效率低、工作环境相对危险,难以快速反映海岸线动态变化的问题,提出一种基于天绘一号卫星多光谱影像的海岸线自动提取与分类方法。所提方法首先结合NDWI指数和水平集模型实现全自动水边线提取,然后采用面向对象分割技术和空间拓扑关系提取临海地物基元,最后结合遥感影像海岸线解译标志建立决策树完成临海基元的分类与海岸线提取。实验表明,海岸线提取结果能够作为1∶2.5万海图测制与更新的参考数据。

Aimiiig at the problem that the conventional artificial shoreline survey method consumes human and material resources,is inefficient and the working environment is relatively dangerous,it is difficult to reflect the dynamic changes of the shoreline rapidly.An automatic shoreline extraction and classification method based on the multi-spectral image of TH-1 is proposed. The proposed method first combines NDWI index and horizontal set model to realize automatic water edge extraction, then uses object-oriented segmentation technology and spatial topology relation to extract seaward ground element, and finally combines remote sensing image to establish decision tree to finish seaward element classification and shoreline extraction. The experimental results show that the coastline extraction can be used as reference data for 1-25,000 nautical chart measurement and updating.