基于纹理信息CART决策树的林芝县森林植被面向对象分类

发布时间：2018-01-04 16:18

本文关键词：基于纹理信息CART决策树的林芝县森林植被面向对象分类　出处：《遥感技术与应用》2017年02期 　论文类型：期刊论文

【摘要】：以西藏自治区林芝县的Landsat-8影像、地形图为信息源,结合样地调查数据及森林资源二类调查数据,研究基于纹理信息的CART决策树的面向对象分类对研究区内的森林地物类别进行提取,分类的总体精度和Kappa系数分别为82.53%和0.768,相较于不利用纹理信息的决策树分类和基于最大似然分类法的研究区地物类别的提取总体精度均高近10%,Kappa系数分别高0.12和0.111。结果表明:基于纹理信息的CART决策树面向对象分类方法对研究区Landsat-8影像进行植被类型提取,分类结果较好,能够满足研究要求。
[Abstract]:The Landsat-8 image of Linzhi County of Tibet autonomous region, topography as a source of information, combined with the survey data of two sample survey data and forest resources, the research object oriented CART decision tree classification of texture information for the study area of forest land cover categories were extracted based on the classification, the overall accuracy and Kappa coefficient were 82.53% and 0.768 and compared to the decision tree using texture information classification and extraction accuracy of the overall study area features the maximum likelihood classification method based on the high of nearly 10%, Kappa coefficient was 0.12 higher and 0.111. results show that the image texture information of decision tree CART object-oriented classification method of Landsat-8 based on vegetation extraction, classification the result is good, can meet the research requirements.

【作者单位】：中国林业科学研究院资源信息研究所;西藏自治区林业调查规划研究院;
【基金】：国家科技基础性工作专项“中国森林植被调查”(2013FY111600)
【分类号】：TP751
【正文快照】： 2.西藏自治区林业调查规划研究院,西藏拉萨850000)1引言森林是地球上最大的陆地生态系统,具有森林资源、调节气候、水源涵养、土地沙漠化防治、生物多样性保护和生态平衡维护等价值,在全球生物圈中具有重要影响[1]。遥感技术是远距离在不接触目标物体的情况下,通过接受目标物

【相似文献】