论文部分内容阅读
为了实现水稻品种的快速鉴别,避免水稻品种混杂,利用电子鼻对来自同一产地不同水稻品种进行测试,获取有效信息。对获取的信息提取平均微分值和面积斜率比两种特征。采用主成分分析、Fisher判别分析及BP神经网络3种模式识别方法进行水稻品种的判别,并对3种识别方法的结果进行比较分析。结果表明:不同种类的水稻品种可以被区分开来,但BP神经网络分类效果最好,Fisher判别分析效果次之,PCA分类效果最差。因此,结合合适的特征提取方法及模式识别方法,有可能实现一种基于电子鼻技术的对不同水稻品种鉴别的简单、有效的方法。
In order to achieve rapid identification of rice varieties and to avoid hybrid rice varieties, electronic rice was used to test different rice varieties from the same place of production for effective information. The extracted information is extracted from the average differential value and area slope ratio of two characteristics. The principal components analysis, Fisher discriminant analysis and BP neural network were used to discriminate the rice varieties. The results of the three identification methods were compared and analyzed. The results showed that different varieties of rice could be distinguished, but BP neural network had the best classification effect, Fisher discriminant analysis followed by PCA classification. Therefore, it is possible to realize a simple and effective method for identifying different rice varieties based on electronic nose technology, combining with the appropriate feature extraction method and pattern recognition method.