基于遥感影像的水边线提取方法综述

Review of waterline extraction methods based on remote sensing images

海岸线是体现自然因素和人类活动对沿海环境影响的关键,在沿海资源开发、海岸带综合管理等领域具有重要的应用价值。基于遥感技术的海岸线提取步骤一般是先通过遥感影像提取瞬时水边线,然后对其进行潮位校正。本文针对海岸线分析的重要内容——水边线提取进行综述,主要介绍了遥感影像提取水边线的传统和人工智能两类方法,其中传统的提取方法包括阈值分割法、区域分割法、边缘检测法和面向对象法,人工智能的提取方法包括传统机器学习法和深度学习法。在总结和分析各种具体水边线提取方法的特点和优缺点基础上,展望了未来水边线提取的研究或发展方向。本文认为未来利用遥感影像进行水边线提取可以重点关注基于高分辨率影像的水边线精细提取、基于深度学习方法的水边线全自动提取、不同海岸水边线多类别提取以及基于无人机遥感的水边线提取应用等几个方向。

The coastline is the key to reflecting the impact of natural factors and human activities on the coastal environment,and holds significant practical value in fields such as coastal resource development and coastal zone comprehensive management.The process of extracting the coastline based on remote sensing technology typically involves first extracting the instantaneous water edge using remote sensing imagery and then correcting it for tidal variations.This article summarizes the waterline extraction methods,which are critical for the coastline analysis,mainly introducing the traditional and artificial intelligence methods for extracting waterline from remote sensing images.The traditional methods include threshold segmentation,region segmentation,edge detection,and object-oriented methods.The artificial intelligence methods include traditional machine learning and deep learning.On the basis of summarizing and analyzing the characteristics,advantages and disadvantages of various specific waterline extraction methods,this article looks forward to the future research directions of waterline extraction.This article believes that in the future,the use of remote sensing images for waterline extraction can focus on fine extraction of waterline based on high-resolution images,fully automatic extraction of waterline based on deep learning methods,multi category extraction of waterline from different coasts,and application of waterline extraction based on drone remote sensing.