测绘通报 ›› 2017, Vol. 0 ›› Issue (10): 22-28.doi: 10.13474/j.cnki.11-2246.2017.0310

• 学术研究 • 上一篇    下一篇

GlobeLand30湿地细化分类研究

陈炜1,2, 陈利军2, 陈军2, 陈浩3, 周晓光1, 谢波3   

  1. 1. 中南大学地球科学与信息物理学院, 湖南 长沙 410083;
    2. 国家基础地理信息中心, 北京 100830;
    3. 湖南科技大学地理空间信息技术国家地方联合工程实验室, 湖南 湘潭 411201
  • 收稿日期:2017-01-05 修回日期:2017-05-28 出版日期:2017-10-25 发布日期:2017-11-07
  • 作者简介:陈炜(1991-),男,硕士,主要研究方向为摄影测量与遥感。E-mail:365266097@qq.com
  • 基金资助:
    国家高技术研究发展计划(863计划)(2013AA122802)

Research on Wetland Sub-classification from GlobeLand30

CHEN Wei1,2, CHEN Lijun2, CHEN Jun2, CHEN Hao3, ZHOU Xiaoguang1, XIE Bo3   

  1. 1. School of Geosciences and Info-physics, Central South University, Changsha 410083, China;
    2. National Geomatics Center of China, Beijing 100830, China;
    3. National-local Joint Engineering Laboratory of Geo-spatial Information Technology, Hunan University of Science and Technology, Xiangtan 411201, China
  • Received:2017-01-05 Revised:2017-05-28 Online:2017-10-25 Published:2017-11-07

摘要: 基于30 m地表覆盖数据产品完成湿地精细化分类,能够更好地满足当前较高分辨率及较详尽全球湿地数据的应用需求。本文在深入分析湿地分类体系与细化方法的基础上,提出以湿地细化类别的定义、多元知识的分层分类、亚类数据精细化提取为主线的总体研究思路,制定了基于先验知识的对象系统筛选、基于森林数据的同位像元提取、基于最佳阈值的极大似然掩膜的主体分类方法,并应用于数据生产实践获得8个亚类信息。该方法克服了常规手段普遍存在的周期长、效率低等弊端,实现了全球较高分辨率湿地亚类数据的快速精确制图,总体分类精度达82.6%,对地理世情及其他地表覆盖研究具有借鉴意义。

关键词: 湿地细化, 分层分类, 全球地表覆盖数据(GlobeLand30), 亚类信息

Abstract: Wetland data of the Globaland30 have been classified into 8 subclasses by the improved method, which could meet the application requirements of high resolution and detailed global wetland data. This paper proposed the research framework of wetland classification, formulated the classification method, and analyzed the 8 subclasses at a global scale. The framework was composed of the definition of wetland subclass, the hierarchical classification of multiple knowledge and methods of sub date extraction. The data have been classified by the object filtering of prior knowledge, the parity pixel extraction of forest data and mask method of the maximum likelihood algorithm. The method overcomed the shortcomings of the conventional methods, such as long cycle and low efficiency, and realized the fast and accurate mapping of the global high resolution wetland subclass data,the accuracy for general wetland subclass is 82.6%. This research would be useful for monitoring and studying the land cover information in the further.

Key words: wetland sub-classification, hierarchical classification, global land cover data (GlobeLand30), subclass information

中图分类号: