摘要
森林类型遥感影像自动制图在森林资源调查中有重要应用,本文以广西壮族自治区金秀县为研究区,基于多时相的Landsat-8数据,采用面向对象的决策树分类方法,对研究区森林类型自动识别进行了研究,分类结果表明:1)单一时相影像森林分类精度中,生长季前期最高,生长季末期次之,非生长季最冷月最低;2)结合生长季与非生长季的多时相影像森林类型自动识别精度较单一时相影像显著提高。
The automatic identification of forest type remote sensing image is an important application in forest resource survey. In this paper, Jinxiu County of Guangxi province is a research area. Based on the multi-temporal landsat8 data, the object-oriented decision tree classification method is used to identify the forest type in the study area.1） The accuracy of classification of single temporal image is the highest in the early growth season, followed by the end of the growth season and the lowest in the non-growing season. 2） The accuracy of automatic recognition of forest type in the combination of the growth season and the non-growing season is more obvious than that of single time image.
作者
李振
胡慧萍
杨敏华
陈鸣
LI Zhen;HU Huiping;YANG Minhua;CHEN Ming(School of Geosciences and Info-Physics,Central South University,Changsha 410083,China;Key Laboratory for Ecotourism of Hunan Province,Changsha 427000,China)
出处
《测绘与空间地理信息》
2018年第9期147-149,共3页
Geomatics & Spatial Information Technology
