| 摘要: |
| 针对海岸线区域地形复杂和卫星遥感影像分辨率的不足，精度难以满足大比例尺成图要求，以及常规解译方法的局限性，选取青岛小岛湾海岸线为研究区，以无人机（UAV）遥感影像为基础数据，提出一种面向对象的海岸线提取方法，结合现场实测验证，开展了人工海岸线和砂质海岸线识别的应用实验。结果表明：人工海岸线和砂质海岸线概率边缘指数（PRI）分别为0.97和0.88，边缘定位误差（BDE）分别为4.33和2.84，提取的人工海岸线和砂质海岸线与实测海岸线结果整体上匹配较好，仅在局部细微处存在微小差异。本文提出的方法可快速有效地获取海岸线信息，其精度能够满足海岸线动态变化监测的需求，可在海岸线资源管理中推广应用。 |
| 关键词: 无人机(UAV)遥感 面向对象 半自动化 海岸线提取 |
| DOI：10.11759/hykx20200509001 |
| 分类号:P76 |
| 基金项目:海洋公益性科研专项（201405028-4） |
|
| Research on object-oriented method of extracting coastline from unmanned aerial vehicle remote sensing image |
|
MA De-ming1,2, LIU Yan-xiong1, JIN Yong-de1, CHENG Bao-quan3
|
|
1.The First Institute of Oceanography, Ministry of Natural Resources, Qingdao 266061, China;2.College of Environmental Sciece and Engineering, Ocean University of China, Qingdao 266100, China;3.Qingjian Group Co., Ltd., Qingdao 266071, China
|
| Abstract: |
| The coastline is one of the important signs of the coastal zone, and also the baseline and “starting line” of the marine development strategy. It not only marks the boundary between land and sea, but also contains rich environmental information, and has an important role in indicating the coastal beach utilization and the rise and fall of wetland ecosystem. As a precious natural resource, coastline plays an important role in coastal economic development. How to quickly and accurately extract the information of coastline location and its dynamic changes is an urgent problem for natural resources authorities and researchers. In this paper, the coastline of Qingdao small Island Bay was taken as the research object, and the unmanned aerial vehicle remote sensing image was taken as the basic data. An object-oriented coastline extraction method was proposed. Combined with the field test, the application experiment of artificial coastline and sandy coastline identification was carried out. The results show that the probability edge index PRI (Probabilistic Rand Index) of artificial coastline and sandy coastline are 0.97 and 0.88, respectively, and the edge positioning error BDE (Boundary Displacement Error) is 4.33 and 2.84, respectively. The extracted artificial coastline and sandy coastline are basically similar to the measured coastline results. There are only slight differences in local fineness, and the accuracy of coastline identification is high. It is proved that the method can quickly and effectively acquire the coastline information, and its accuracy can meet the needs of coastline dynamic change monitoring, and can be promoted and applied in coastline resource management. |
| Key words: unmanned aerial vehicle remote sensing object-oriented semi-automatic coastline extraction |
