| 基于LBP算法的全天空红外云图的分类研究 |
| |
| 引用本文： | 孙学金,刘磊,高太长,赵世军.基于LBP算法的全天空红外云图的分类研究[J].南京气象学院学报,2009,32(4):490-497. |
| |
| 作者姓名： | 孙学金 刘磊 高太长 赵世军 |
| |
| 作者单位： | 1. 解放军理工大学气象学院,江苏,南京,211101;北京大学物理学院,北京,100871
2. 解放军理工大学气象学院,江苏,南京,211101 |
| |
| 基金项目： | 公益性行业(气象)科研专项(GYHY |
| |
| 摘 要： | 利用局部二值模式（Local Binary Patterns，LBP），结合局部云图的方差信息（VAR），对全天空红外测云系统获得的红外云图进行了分类研究。首先建立了器测云状分类与传统分类的对应关系，其次分析了层状云、积状云、波状云、卷云和晴空5种类型天空的LBP谱和VAR特征，最后对274个样本进行了分类识别，并与人工观测结果比较，结果表明，层状云、积状云、波状云、卷云和晴空的识别正确率分别为100．0％、84．2％、70．3％、64．7％、99．0％，平均正确率达到87．2％。
|
| 关 键 词： | LBP算法 全天空红外测云系统 云分类 |
Classification of Whole Sky Infrared Cloud Image Based on the LBP Operator |
| |
| SUN Xue-jin,LIU Lei,GAO Tai-chang,ZHAO Shi-jun.Classification of Whole Sky Infrared Cloud Image Based on the LBP Operator[J].Journal of Nanjing Institute of Meteorology,2009,32(4):490-497. |
| |
| Authors: | SUN Xue-jin LIU Lei GAO Tai-chang ZHAO Shi-jun |
| |
| Institution: | SUN Xue-jin, LIU Lei , GAO Tai-chang , ZHAO Shi-jun(Institute of Meteorology, PLA University of Science and Technology, Nanjing 211101, China 2. School of Physics,Peking University,Beijing 100871 ,China) |
| |
| Abstract: | This paper proposes a classification method of sky-condition based on whole sky infrared cloud image, where the Local Binary Patterns(LBP) operator and the contrast of local cloud image tex-ture(VAR signal) are used to classify sky conditions. The correspondence relationship between tradi-tional cloud classes and instrument-measured cloud classes is suggested;and the LBP spectra and VAR characteristics for the five classes of stratus, cumulus, undulatus, cirrus clouds and clear sky are ana-lyzed. The classification correct rates for the above five classes in the class-recognition of 274 cloud im-age samples are 100. 0%, 84. 2%, 70. 3%, 64. 7%, and 99.0%, respectively, with an average of 87.2%. |
| |
| Keywords: | Local Binary Patterns whole sky infrared cloud measuring system sky-condition classifica-tion |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
