eCognition在西藏全区草原普查遥感影像信息快速提取中的应用
详细信息 查看官网全文
- 英文论文题名：Application of eCognition to rapidly extract the information from remote sensing image of grassland general investigation in Tibet region
- 论文作者：陈喜梅 ; 王庆国 ; 王一斐 ; 王丹斓
- 英文论文作者：CHEN Xi-mei ; WANG Qing-guo ; WANG Yi-fei ; WANG Dan-lan ; Institute of Grassland Surveying and Planning ; College of Grassland ; Resources and Environment ; Inner Mongolia Agricultural University
- 年：2016
- 作者机构：内蒙古自治区草原勘察规划院;内蒙古农业大学草原与资源环境学院;
- 论文关键词：草原普查 ; eCognition ; 面向对象 ; 类型分类
- 英文论文关键词：grassland general investigation ; e Cognition ; object-oriented ; type classification
- 会议召开时间：2016-08-19
- 会议录名称：第四届中国草业大会论文集
- 语种：中文
- 分类号：TP751
- 学会代码：ZXMY
- 会议名称：第四届中国草业大会
- 会议地点：中国青海西宁
- 主办单位：中国畜牧业协会草业分会
- 学会名称：中国畜牧业协会
- 页数：8
- 文件大小：1196k
- 原文格式：D
- 会议级别：全国
摘要
以西藏全区草原作为研究区域,利用eCognition软件平台,运用面向对象的多尺度分割方法,对研究区中高分辨率Spot、TM遥感影像探索性的进行草原类型分类。该方法综合利用了光谱、形状与纹理等特征,快速、有效地进行草原类型的信息分类提取,取得了良好的效果,总体精度达到93.69%,草原类型分类精度达到了90.71%。e Cognition面向对象分类技术首次在西藏草原普查中进行了应用,丰富了草原普查的技术手段与方法,是一次比较成功的分类应用探索。
The survey region of this study was Tibet region grassland. The grassland types was classified by remote sensing image of high resolution Spot and TM using e Cognition software platform and applying the object oriented multiscale segmentation method. This method synthetically utilized the characteristic of spectrum, shape, lines, etc. and extracted rapidly and effectively the information classification of grassland types. The overall accuracy reached 93.69% and the accuracy of grassland type classification reached 90.71%, which obtained good effect. e Cognition which is object-oriented classification methods was used in grassland general investigation in Tibet region for the first time. Technological means and methods of grassland general investigation were enriched, which was successful exploration of grassland classification application.
The survey region of this study was Tibet region grassland. The grassland types was classified by remote sensing image of high resolution Spot and TM using e Cognition software platform and applying the object oriented multiscale segmentation method. This method synthetically utilized the characteristic of spectrum, shape, lines, etc. and extracted rapidly and effectively the information classification of grassland types. The overall accuracy reached 93.69% and the accuracy of grassland type classification reached 90.71%, which obtained good effect. e Cognition which is object-oriented classification methods was used in grassland general investigation in Tibet region for the first time. Technological means and methods of grassland general investigation were enriched, which was successful exploration of grassland classification application.
引文