基于Android平台的地基云状识别系统  

Ground Nephogram Recognition System Based on Android Platform

作　　者：张之正 冯径 徐志康 武亚军 ZHANG Zhi-zheng;FENG Jing;XU Zhi-kang;WU Ya-jun(College of Meteorology and Oceanography,National University of Defence Technology,Nanjing 211101 China;32135 Troops,the Chinese People's Liberation Army,Datong 037000 China)

机构地区：[1]国防科技大学气象海洋学院,江苏南京211101 [2]中国人民解放军32135部队,山西大同037000

出　　处：《自动化技术与应用》2019年第11期127-131,146,共6页Techniques of Automation and Applications

基　　金：国家自然科学基金资助项目（编号61371119）

摘　　要：随着社会对于气象关注程度的提高以及智能手机的广泛使用,人们对于气象服务的要求更加多样化。为了丰富气象信息服务手段,开发了一款基于Android平台的地基云状识别系统。该系统以Dense_SIFT作为云状的特征描述子,利用词袋模型创建视觉词汇本,选择支持向量机(SVM)作为分类器。系统包括图像获取、图像预处理、特征提取、分类识别、结果展示5个模块。实验表明,该系统能够有效地对不同类别的地基云图进行分类并展示,具有较好的识别效果,达到了预期目标。With the increase of social attention to meteorology and the popularization of smart phones,people have more diversified demands on weather information.In order to enrich meteorological service approaches,a ground nephogram recognition system based on android platform is proposed in this paper.Dense_SIFT is used for describing the feather of ground and support vector machine(SVM)is selected as the classifier.The system includes image acquisition,image processing,feature extraction,classification and result display.The experiment shows that the system can effectively classify and display different types of ground nephogram,which has a good recognition effect and achieves the expected goals.

关 键 词：ANDROID 特征提取 云状识别 支持向量机 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]