首页 | 本学科首页   官方微博 | 高级检索  
     

基于概率神经网络的粮食早期霉变识别研究
引用本文:吴莉莉,黄品高,惠国华.基于概率神经网络的粮食早期霉变识别研究[J].计算机应用与软件,2011(9).
作者姓名:吴莉莉  黄品高  惠国华
作者单位:河南农业大学理学院;桂林电子科技大学教学实践部;浙江工商大学食品与生物工程学院;
基金项目:郑州市科技攻关项目(083SGYG241232)
摘    要:提出了一种基于概率神经网络的粮食早期霉变识别方法。实验中电子鼻系统采集了4种粮食作物及霉变数据共8类,对这些数据样本进行特征提取,得到了64组训练数据和48组测试数据。利用概率神经网络对特征数据进行分类识别,识别率为93.75%。实验结果表明,该方法对粮食作物种类及其早期霉变的识别是行之有效的。

关 键 词:早期霉变  概率神经网络  电子鼻  特征提取  

ON IDENTIFYING EARLY GRAINS MILDEWING BASED ON PROBABILISTIC NEURAL NETWORK
Wu Lili Huang Pin'gao Hui Guohua.ON IDENTIFYING EARLY GRAINS MILDEWING BASED ON PROBABILISTIC NEURAL NETWORK[J].Computer Applications and Software,2011(9).
Authors:Wu Lili Huang Pin'gao Hui Guohua
Affiliation:Wu Lili1 Huang Pin'gao2 Hui Guohua3 1(College of Sciences,Henan Agricultural University,Zhengzhou 450002,Henan,China) 2(Department of Teaching and Practice,Guilin University of Electronic Technology,Guilin 541004,Guangxi,China) 3(College of Food Science and Biotechnology,Zhejiang Gongshang University,Hangzhou 310035,Zhejiang,China)
Abstract:A method of recognising early grains mildewing based on probabilistic neural network is presented in this paper.In the experiment,4 sets of grains and their early mildewing data totalling 8 kinds of data were collected by the electronic nose system,and the feature of these data sample were extracted,there were 64 groups training data and 48 groups testing data obtained.The feature data were classified and recognised using probabilistic neural network,the recognition rate was 93.75%.Experimental results show...
Keywords:Early mildewing Probabilistic neural networks Electronic nose Feature extraction  
本文献已被 CNKI 等数据库收录!
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