热带海洋学报 ›› 2020, Vol. 39 ›› Issue (4): 116-129.doi: 10.11978/2019100

• 海洋调查与观测 • 上一篇    

基于高分辨率遥感的珊瑚礁地貌单元体系构建和分类方法—以8波段Worldview-2影像为例

董娟1(), 任广波1(), 胡亚斌1,2, 逄今朝1,3, 马毅1   

  1. 1.自然资源部第一海洋研究所, 山东 青岛 266061
    2.大连海事大学信息科学技术学院, 辽宁 大连 116026
    3.中国石油大学(华东), 山东 青岛 266580
  • 收稿日期:2019-10-12 修回日期:2020-02-02 出版日期:2020-07-20 发布日期:2020-07-27
  • 通讯作者: 任广波
  • 作者简介:董娟(1995—), 女, 硕士研究生, 主要从事珊瑚礁遥感监测研究。Email: dongj_0829@163.com
  • 基金资助:
    国家自然科学基金项目(51839002)

Construction and classification of coral reef geomorphic unit system based on high-resolution remote sensing: using 8-band Worldview-2 Image as an example

Juan DONG1(), Guangbo REN1(), Yabin HU1,2, Jinzhao PANG1,3, Yi MA1   

  1. 1. First Institute of Oceanography, Ministry of Natural Resources, Qingdao 266061, China
    2. Information Science and Technology College, Dalian Maritime University, Dalian 116026, China
    3. China University of Petroleum, Qingdao 266580, China
  • Received:2019-10-12 Revised:2020-02-02 Online:2020-07-20 Published:2020-07-27
  • Contact: Guangbo REN
  • Supported by:
    Foundation item: National Natural Science Foundation of China(51839002)

摘要:

珊瑚礁生态系统是全球初级生产力最高的生态系统之一, 在维持海洋生物多样性、防浪固滩、资源供给等方面发挥着巨大作用。珊瑚礁遥感地貌分类体系是珊瑚礁保护、管理及可持续发展的必要基础, 但目前还未有结合珊瑚覆盖度进行分类的体系。本文基于WorldView-2高分辨率遥感影像, 以中国南海西沙群岛七连屿北部赵述岛和西沙洲所在礁盘作为研究区, 结合珊瑚覆盖度、区域地貌成分和水动力条件等指标, 建立了既适用于遥感监测又与珊瑚生存状况相关联的珊瑚礁地貌单元分类体系。同时, 利用面向对象的支持向量机(Support Vector Machine, SVM)和随机森林(Random Forest, RF)分类方法进行珊瑚礁地貌单元的信息提取, 并对分类结果进行精度评价。结果表明, SVM和RF两种分类方法均能较好地提取出珊瑚礁地貌单元, 分类精度分别为87.59%和79.81%。针对分类过程中出现的错分、漏分问题, 结合珊瑚礁成因和分布规律对分类结果进行修正, 修正后分类提取的精度达到91.3%, Kappa系数为0.9041, 表明本文构建的珊瑚礁地貌单元分类体系在一定程度上能满足当前珊瑚岛礁信息提取的需要。

关键词: 珊瑚礁, 地貌单元, 分类体系, 遥感, 面向对象

Abstract:

Coral reef ecosystem is one of the ecosystems with the highest primary productivity in the world. It plays an important role in maintaining marine biodiversity, preventing waves and consolidating beaches, and it is great resources for mankind. At the same time, the remote sensing geomorphological classification system of coral reefs is the necessary basis for the protection, management and sustainable development of coral reefs. However, there is no corresponding classification system taking coral coverage into account. Based on WorldView-2 high-resolution remote sensing image, we selected the reef of Zhaoshu Dao and Xisha Zhou in the north of Qilianyu Island, Xisha Islands, South China Sea as our study area. In this paper, coral coverage, regional geomorphic composition and hydrodynamic conditions are used to establish a coral reef geomorphic unit classification system suitable for remote sensing monitoring and are related to the survival of coral reefs. The method of object-oriented support vector machine (SVM) and random forest (RF) were used to extract the coral reef geomorphic units, and the accuracy of the extraction results was evaluated. The results showed that both SVM and RF classification methods can extract reef geomorphic units, and the classification accuracy was 87.59% and 79.81%, respectively. Aiming at the problem of misclassification and omission in the process of classification, we modified the classification results according to the causes and distributions of coral reefs. And the accuracy of modified classification was 91.3%, and the kappa coefficient was 0.9041. They showed that the coral reef geomorphic unit classification system constructed in this paper can meet the need of information extraction of coral reefs to some extent.

Key words: coral reef, geomorphology, classification scheme, remote sensing, object oriented

中图分类号: 

  • P737.2