首页 | 本学科首页   官方微博 | 高级检索  
     

叶片图像特征提取与识别技术的研究
引用本文:王晓峰,黄德双,杜吉祥,张国军.叶片图像特征提取与识别技术的研究[J].计算机工程与应用,2006,42(3):190-193.
作者姓名:王晓峰  黄德双  杜吉祥  张国军
作者单位:1. 中国科学院合肥智能机械研究所,合肥,230031;合肥学院计算机科学与技术系,合肥,230022
2. 中国科学院合肥智能机械研究所,合肥,230031
摘    要:文章介绍了一种基于叶片图像的形状特征对叶片进行识别的方法。首先对叶片图像进行预处理并且提取出叶片的轮廓,然后利用轮廓计算得到叶片的矩形度、圆形度、偏心率等8项几何特征和7个图像不变矩,同时在文中提出了一种新的移动中心超球分类器,利用其对得到的形状特征进行分类从而实现了对20多种植物叶片的快速识别,并且平均识别率达到了92%。

关 键 词:叶片图像  图像预处理  特征提取  移动中心超球分类器
文章编号:1002-8331-(2006)03-0190-04
收稿时间:2005-04
修稿时间:2005-04

Feature Extraction and Recognition for Leaf Images
Wang Xiaofeng,Huang Deshuang,Du Jixiang,Zhang Guojun.Feature Extraction and Recognition for Leaf Images[J].Computer Engineering and Applications,2006,42(3):190-193.
Authors:Wang Xiaofeng  Huang Deshuang  Du Jixiang  Zhang Guojun
Affiliation:1.Institute of Intelligent Machines,Chinese Academy of Seienees,Hefei 230031;2.Department of Computer Science and Teehnology,Hefei University,Hefei 230022
Abstract:This paper introduces a method of recognizing leaf images based on shape features.Firstly,we apply image pre-processing techniques to get the contours of the leaves.Then we extract eight geometric features including rectangularity,circularity,eccentricity,etc,and seven moment invariants from leaves' contours for classification.Finally we propose a new moving menter hypersphere classifier to address these shape features.As a result there are more than 20 kinds of plant leaves successfully recognized.The average correct recognition rate is up to 92 percent.
Keywords:leaf images  image pre-processing  feature extraction  moving center hypersphere classifier
本文献已被 CNKI 维普 万方数据 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