首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于统计方法的地基器测云天识别
引用本文:秦健,刘磊,高太长,孙学金.基于统计方法的地基器测云天识别[J].气象科学,2011,31(5):604-612.
作者姓名:秦健  刘磊  高太长  孙学金
作者单位:1. 解放军理工大学气象学院,南京211101;95129部队气象台,河南开封475003
2. 解放军理工大学气象学院,南京,211101
基金项目:公益性行业(气象)科研专项(GYHY200806030)
摘    要:本文从图像处理的角度,基于统计方法对地基全天空类型识别进行了研究。通过对2007年1月的397组观测数据进行统计分析,确定了5种天空类型识别的判别阈值,实现了地基全天空类型的有效识别。结果表明,样本试验的分类结果为81.36%,普适性检验的结果为79.51%。

关 键 词:地基云分类  统计方法  天空类型识别
收稿时间:4/6/2010 12:00:00 AM
修稿时间:2010/10/2 0:00:00

Study on the recognition of whole sky condition types based on the statistical method
QIN Jian,LIU Lei,GAO Taichang and SUN Xuejin.Study on the recognition of whole sky condition types based on the statistical method[J].Scientia Meteorologica Sinica,2011,31(5):604-612.
Authors:QIN Jian  LIU Lei  GAO Taichang and SUN Xuejin
Institution:QIN Jian1,2 LIU Lei1 GAO Taichang1 SUN Xuejin1(1 Institute of Meteorology,PLA University of Science and Technology,Nanjing 21110,CHina,2 Observatory of PLA95129,Henan Kaifeng 475003,China)
Abstract:The recognition of ground-based whole-sky condition types has been studied by the statistical method from the viewpoint of digital image processing.Based on the statistical analysis of the data during January 2007,this paper gives the thresholds of five sky condition types separately,and then the statistical method of the recognition of whole-sky condition types has been established.The results of the experimentation show that the average index of agreement is about 81.36% for the sample cloud maps and 79.8...
Keywords:Statistical method  Ground-based cloud classification  Recognition of whole-sky condition types  
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《气象科学》浏览原始摘要信息
点击此处可从《气象科学》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号