测绘通报 ›› 2023, Vol. 0 ›› Issue (8): 24-28.doi: 10.13474/j.cnki.11-2246.2023.0226

• 学术研究 • 上一篇    下一篇

无人机遥感影像浒苔信息提取方法

麻德明1, 张秀林2, 田梓文1, 王佳新3   

  1. 1. 自然资源部第一海洋研究所, 山东 青岛 266061;
    2. 青岛市黄岛区海洋发展局, 山东 青岛 266400;
    3. 青岛农高陡崖子水务有限公司, 山东 青岛 266555
  • 收稿日期:2023-03-01 发布日期:2023-09-01
  • 通讯作者: 张秀林。E-mail:bjqdlele@163.com
  • 作者简介:麻德明(1982-),男,博士,高级工程师,主要研究方向为海岛海岸带环境遥感。E-mail:demingma@fio.org.cn
  • 基金资助:
    海洋公益性科研专项(201405028-4)

Information extraction method of Enteromorpha prolifera based on UAV remote sensing image

MA Deming1, ZHANG Xiulin2, TIAN Ziwen1, WANG Jiaxin3   

  1. 1. First Institute of Oceanography, Ministry of Natural Resources, Qingdao 266061, China;
    2. Qingdao Municipal Huangdao District Marine Development Bureau, Qingdao 266400, China;
    3. Qingdao Nonggao Douyazi Water Service Co., Ltd., Qingdao 266555, China
  • Received:2023-03-01 Published:2023-09-01

摘要: 浒苔作为一种典型的海洋污染,不仅造成严重的生态环境问题,而且对沿海经济发展产生重大的影响,如何快速、准确地获取浒苔的位置、边界范围及其动态变化信息,是自然资源主管部门和科研工作者关注的焦点问题。本文选取大管岛及其附近海域为研究区,以无人机遥感影像为基础数据,提出了一套面向多场景浒苔的边界范围智能解译方法,结合人工解译验证,分别开展了沿岸和海洋面中浒苔识别的应用实例。结果表明,海岸和海洋面浒苔提取的总体精度分别为96.75%和98.13%,Kappa系数分别为0.72和0.71,提取的浒苔边界范围与人工解译结果整体上匹配较好。本文方法可准确有效地获取浒苔边界范围信息,其精度能够满足浒苔信息识别的需求,可为高精度浒苔跟踪、精细化的灾害预警与防控提供数据参考和技术支撑。

关键词: 无人机遥感, 浒苔, 智能解译, 提取方法, 精度评估

Abstract: Enteromorpha prolifera, as a typical marine pollution, not only causes serious ecological and environmental problems, but also has a significant impact on coastal economic development. How to quickly and accurately obtain the location, boundary range and dynamic change information of Enteromorpha prolifera is the focus of natural resource authorities and researchers. In this paper, Daguan Island and its adjacent waters are selected as the study area, and a set of intelligent interpretation methods for the boundary range of Enteromorpha prolifera in multiple scenes is proposed based on UAV remote sensing images. At the same time, combined with the manual interpretation and verification, the application examples of Enteromorpha prolifera identification in coastal and marine areas are carried out respectively. The results show that the overall accuracy of the extraction of Enteromorpha prolifera from coastal and marine surfaces is 96.75% and 98.13%, respectively, and the Kappa coefficient is 0.72 and 0.71, respectively. The boundary range of Enteromorpha prolifera extracted matches well with the result of manual interpretation. The method proposed in this paper can quickly and effectively obtain the boundary range information of Enteromorpha prolifera, and its accuracy can meet the requirements of Enteromorpha prolifera information identification, and can provide data reference and technical support for high-precision tracking, refined disaster early warning and prevention and control.

Key words: unmanned aerial vehicle remote sensing, Enteromorpha prolifera, intelligent interpretation, extraction method, accuracy evaluation

中图分类号: