摘要
植被图是表示多种植被或植物群落的空间分布规律及其生态环境的地图,它是生物多样性保护、生态研究、自然资源管理和生态恢复的重要依据.目前,中国植被专题资源信息的本底数据《中华人民共和国植被图(1:1000000)》从开始绘制至今已将近40年,中国的植被分布格局已经发生了极大的变化,《中华人民共和国植被图(1:1000000)》已存在植被斑块的类别和边界与现实不符等问题,中国的植被分布本底数据亟待更新.如今卫星遥感技术的发展为实现大面积区域的植被制图提供了一种实用且经济的手段.本文综述了国家尺度植被图的制图方法和卫星遥感技术在植被分类制图上的进展,并以此为基础,探讨中国新一代1:50万植被图的遥感制图方法.新一代1:50万中国植被图的绘制通过众源采集结合专家鉴定的方式获取海量植被类型样本,基于多源遥感数据,以植被斑块为对象,采用深度学习的方式实现遥感植被分类,并基于自主构建的植被在线平台,借助于全国各地的植被生态学家的专业知识实现对制图结果的校订和更新.
Vegetation maps serve as the key source information for ecological studies, biodiversity conservation, and vegetation management and restoration. The latest version of the Vegetation Map of China(1:1,000,000) was generated in the 1980s. Since then, the vegetation distribution pattern of China has changed dramatically during these 40 years. Classification errors and time lag have limited the applications of Vegetation Map of China(1:1,000,000), and it is in great demand to make the new generation of national vegetation map to fulfill the needs of ecological studies and government policy making. The development of satellite remote sensing technology provides a practical and economical approach to achieve vegetation mapping in large scale. In this article, we reviewed methods of vegetation mapping at national scale and the progress of satellite remote sensing technology on vegetation classification and mapping, and summarized the current bottleneck in vegetation mapping from satellite images. Further, we introduced a vegetation mapping strategy through the combination of crowdsource sample collection, object-based segmentation, and deep learning techinique from multi-source data. Over 50 taxonomists across China participated in the validation and calibration process through a self-developed online mapping system.
作者
郭庆华
关宏灿
胡天宇
金时超
苏艳军
王雪静
魏邓杰
马勤
孙千惠
GUO QingHua;GUAN HongCan;HU TianYu;JIN ShiChao;SU YanJun;WANG XueJing;WEI DengJie;MA Qin;SUN QianHui(State Key Laboratory of Vegetation and Environmental Change,Institute of Botany,Chinese Academy of Sciences,Beijing 100093,China;University of Chinese Academy of Sciences,Beijing 100049,China)
出处
《中国科学：生命科学》
CSCD
北大核心
2021年第3期229-241,共13页
Scientia Sinica(Vitae)
基金
中国科学院战略性先导科技专项(A类)“地球大数据科学专项”(批准号:XDA19050401)资助。
关键词
中国植被图
植被制图
卫星遥感
众源采集
深度学习
Vegetation Map of China
vegetation mapping
satellite remote sensing
crowdsource
deep learning
