摘要
在植物的识别问题上,一些相似植物之间的界线不一定很清晰,一些识别带有个人主观色彩。数学识别则是按一定的数学方法,建立数学模型,对样本进行识别,相对而言比较客观。叶片是植物非常明显的一种外部形态特征。选取叶形相似的苦槠和青冈栎的叶片作为研究对象,针对这两种叶片的识别问题,以叶片长度及叶片1/2处宽度作为定量指标,建立F isher判别函数并对样本叶片进行判别,结果显示:样本数据的回判率为94%,扩大样本的回判率为92%。
To plant-identification,some indentification of the similar plants may be not clear,so that the identification may be subjective.So,by corresponding mathematical methods to identify samples,the relatively objective results can be get.Leaf is one of the most visible morphological features of plant.Castanopsis sclerophylla and Cyclobalanopsis glauca's leaves with similar shapes were taken as the research samples in this paper.Based on the two quantitative indexes of leaf length and leaf width at half-length,the Fisher discriminant function was established and used to make discriminant analysis for the sample leaves.It is proved that the function is practical with discrimination accuracies of 94% to the sample leaves and 92% to the expanded sample leaves.
出处
《河南科技大学学报（自然科学版）》
CAS
北大核心
2010年第6期62-65,共4页
Journal of Henan University of Science And Technology:Natural Science
基金
国家林业局重点项目(2001-5)
国家科技支撑项目(2006BAD03A0702/wb01)
关键词
Fisher判别分析法
苦槠
青冈栎
叶片
定量分析
Fisher discriminant analysis
Castanopsis sclerophylla
Cyclobalanopsis glauca
Leaves
Quantitative analysis
