%0 Journal Article
%T Forest Fuel Types Classification and Change Detection by Remote Sensing in Zhangping City<br>漳平市森林可燃物类型变化遥感动态监测
%A CHEN Yunzhi
%A CHEN Chongcheng
%A WANG Xiaoqin
%A TANG Liyu
%A <br>陈芸芝
%A 陈崇成
%A 汪小钦
%A 唐丽玉
%J 地球信息科学
%D 2008
%I 
%X Forest fuel type's spatial distribution is one of the key factors to be considered during forest fire spread and fire-fighting visualization modeling and simulation.Based on review of the status of the study on forest fuel type classification,a classification method considering tree types was put forward.According to the attribute of predominant tree types of 2003 forest stand layers,four forest fuel types,i.e.,bamboo forest,broad-leaved forest,Chinese fir forest,and Masson pine forest and their spatial distribution in Zhangping was acquired firstly.The object-oriented classification technology,which using the techniques of image segmentation to construct objects containing information of size,shape,topography and class hierarchy besides spectral information,and involving procedures of segmentation,construction of class hierarchy,and classification,was then applied to the ASTER images to detect interior and exterior change of each type's.Accuracy assessment of the classification result indicated that by the technology of object-oriented classification,both the spectral features and class-related features were used,and the accuracy was improved in terms of area,reaching 89.3%.Because thematic layers were involved during the image segmentation,the boundaries of objects did not get across the boundaries of each thematic layer,this makes the updating of the existing layers become very convenient.Finally,the vector format data of remote sensing classification result was used to update the original forest stand layer in order to get the available inputs of fire visualization modeling.This study not only provided up-to-date input layers of forest fuels types for fire visualization modeling,but also detected the areas which have been changed without changing the boundaries of the original forest stands.For forest resource managers,it provided target areas for validation by field survey.
%K forest fuel type
%K ASTER image
%K object-oriented classification
%K updating of forest stand<br>森林可燃物类型
%K ASTER影像
%K 面向对象分类
%K 小班图层更新
%U http://www.alljournals.cn/get_abstract_url.aspx?pcid=E62459D214FD64A3C8082E4ED1ABABED5711027BBBDDD35B&cid=DA72A78627FE64EAA572951EA05D274A&jid=366CD77E372F1875C812440223DE7AFB&aid=1BF174B24757562EA7E25BACA5EE3760&yid=67289AFF6305E306&vid=F3090AE9B60B7ED1&iid=B31275AF3241DB2D&journal_id=1560-8999&journal_name=地球信息科学&referenced_num=1&reference_num=12