基于PCA和SVM的植物叶片分类方法研究  

Plant Leaves Classification Based on PCA and SVM

作　　者：张昭[1] 杨民仓[2] 何东健[3] 

机构地区：[1]宝鸡文理学院电子电气工程系,陕西宝鸡721016 [2]宝鸡职业技术学院生物工程系,陕西宝鸡721013 [3]西北农林科技大学机械与电子工程学院,陕西杨凌712100

出　　处：《农机化研究》2013年第11期34-37,41,共5页Journal of Agricultural Mechanization Research

基　　金：国家自然科学基金项目(60975007);宝鸡文理学院重点项目(ZK10157)

摘　　要：为了提高植物叶片识别与分类的正确率,提出了一种基于PCA和SVM的植物叶片识别方法,在对叶片图像进行分割、边缘检测后,提取10个具有旋转、比例、平移不变性的无量纲叶片特征参数,对叶片特征参数进行主成分分析,将前3个主成分作为支持向量机的输入建立模式识别模型。实验结果表明,本算法女贞、木瓜、五角枫、三角枫等4种植物叶片的识别正确率达97.22%,优于直接用特征参数作为模型输入的识别正确率,且算法具有良好的实时性。To improve the correct rate of identification and classification of plant leaves,a model based on SVM and PCA was proposed.After pre-processing the leaves image including segment and edge detection,10 shape characteristics with rotation,scale and shift-invariance are extracted and the top three principal components were extracted as the input of SVM classifiers by principal component analysis（PCA）.The results show that the average accuracy is above 97.22% for four plant leaves of papaya,privet,mono maple and triangular maple,better than using all the 10 parameters as the input of classifiers,and the algorithm has good real-time performance.

关 键 词：叶片图像 分类识别 特征提取 SVM PCA 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]