基于克隆选择算法和K近邻的植物叶片识别方法

开篇：润墨网以专业的文秘视角，为您筛选了一篇基于克隆选择算法和K近邻的植物叶片识别方法范文，如需获取更多写作素材，在线客服老师一对一协助。欢迎您的阅读与分享！

文章编号:10019081(2013)07200905

doi:10.11772/j.issn.10019081.2013.07.2009

摘 要：

针对植物叶片识别中分类器设计和训练识别时间较长的问题，提出了一种基于人工免疫系统下的克隆选择算法和K近邻判别分析(CSA+KNN)的叶片识别方法。进行图像预处理后，通过提取叶片的几何特征和纹理特征得到叶片综CSA+KNN进行植物叶片样本训练，并进行植物叶片识别。在100种植物叶片数据库中进行测试，CSA+KNN法识别率为91.37%。与BP神经网络等方法相比较，实验结果表明了该识别方法的有效性以及较高的训练速率，同时验证了纹理特征在叶片识别中的重要性。CSA+KNN法扩宽了植物叶片的识别方法，可应用于建立数字化植物标本博物馆等领域。

关键字：植物叶片识别；克隆选择算法；人工免疫系统；数字图像分析；几何特征；纹理特征

中图分类号：TP391.413

文献标志码:A

英文标题

Plant leaf recognition method based on clonal selection algorithm and K nearest neighbor

英文作者名

ZHANG Ning, LIU Wenping*

英文地址(

College of Information, Beijing Forestry University, Beijing 100083, China英文摘要)

Abstract:

To decrease the time of classifier design and training, a new method combining the Clonal Selection Algorithm and K Nearest Neighbor (CSA+KNN) was proposed. Having the image preprocessed and getting the comprehensive features information from geometry and texture feature, the CSA+KNN was used to train and classify the plant leaf samples. The plant leaf database with 100 leaf species was applied to test the proposed algorithm, and the recognition accuracy was 91.37%. Compared with other methods, the experimental results demonstrate the efficiency, accuracy and high training speed of the proposed method, and verify the significance of texture features in leaf recognition. CSA+KNN method broadens the field of plant leaf recognition method, and it can be applied to create digitalized plant specimens museum.

To improve the design of classifier and training time, a new method combining the Clonal Selection Algorithm and K Nearest Neighbor (CSA+KNN) was proposed. Having the image preprocessing and getting the comprehensive features information from geometry and texture feature, the CSA+KNN was used to train and test plant leaf samples. The plant leaf database which had 100 leaf species was applied to test the proposed algorithm, and the recognition accuracy was 91.37%. Compared with other methods, the experimental results demonstrate the efficiency, accuracy and higher training rate of the proposed method, and verify the significance of texture features in leaf recognition. CSA+KNN method broadens the field of plant leaf recognition method, and it can be applied to create digitized plant specimens museum.

英文关键词Key words：

plant leaf recognition; Clonal Selection Algorithm (CSA); Artificial Immune System (AIS); digital image analysis; geometry feature; texture feature

0 引言

植物的分类识别，对于维持生态平衡、保护植物与生物多样性、植物鉴别、植物病害鉴定、建立数字化植物标本博物馆等方面有着非常重要的意义。而植物的叶片作为植物识别的主要参照器官，相比其他器官有着方便采集、存活时间长等优点，基于叶片的植物识别是一种有效且简单的方法。相比传统叶片分类方法，基于图像分析的植物叶片识别有着工作效率高、处理数据广泛、不需要复杂的植物学专业知识等优点。

目前，基于图像分析的植物叶片识别方法有着以下三大类：基于关系结构匹配、基于统计学和基于机器学习的叶片识别方法［1］。近年来，国内外学者在基于图像分析的植物叶片识别方面有一定的进展。2007年，Du等［2］提出了一种移动中值中心超球分类器(Move Median Center Hypersphere Classifier，MMC)，提取15个植物叶片特征用于识别20种植物叶片。2009年，中国科学院合肥物质科学研究院提出一种鲁棒的监督流形学习算法进行植物叶片分类，降低了识别算法的计算复杂度［3］。在2010年，阚江明等［4］将纹理特征运用到叶片识别中，实验证明纹理特征可提高植物识别的准确率。Daliri等［5］在2008年运用动态规划算法对瑞士叶片数据库进行识别检索测试，实验证明此算法的表现优良。2010年，Singh等［6］采用基于二叉树结构的多支持向量机(Multiple Support Vector Machine based on Binary Tree，SVMBDT)方法识别32种植物叶片，实验结果表明此识别方法优于概率神经网络(Probabilistic Neural Network，PNN)分类器和傅里叶矩方法。2011年Sixta［7］利用形状上下文内部距离进行叶片识别。2011年，Rossatto等［8］采用体积分形维数和朴素贝叶斯分类进行树叶图像识别。