引用本文:[点击复制]
[点击复制]
【打印本页】【下载PDF全文】 查看/发表评论下载PDF阅读器关闭

←前一篇|后一篇→

过刊浏览    高级检索

本文已被:浏览次   下载 本文二维码信息
码上扫一扫!
基于图像技术的玉米叶部病害识别研究
祁钊,江朝晖,杨春合,刘连忠,饶元
0
(安徽农业大学信息与计算机学院,合肥 230036)
摘要:
针对野外光照条件下玉米叶部病害的图像识别问题,采用Retinex算法进行图像增强,消除光照的不利影响,在R-G灰度空间中运用自动阈值法进行病斑图像分割,提取病斑的颜色、纹理及不变矩特征,并采用主成分分析和支持向量机相结合的方法进行玉米叶片常见病害的分类识别。实验结果显示,小斑病、锈病和弯孢菌叶斑病的总识别精度为90.74%。表明本研究方法在自然光照环境下可获得良好的病害识别效果,具有一定的实用价值。
关键词:  玉米病害  不均匀光照  Retinex算法  主成分分析  支持向量机
DOI:10.13610/j.cnki.1672-352x.20160311.018
投稿时间:2016-01-20
基金项目:安徽省科技攻关项目(1501031102), 安徽省自然科学基金(1508085MF110)和农业部农业物联网技术集成与应用重点实验室开放基金(2015-kf01)共同资助。
Identification of maize leaf diseases based on image technology
QI Zhao,JIANG Zhaohui,YANG Chunhe,LIU Lianzhong,RAO Yuan
(School of Information and Computer, Anhui Agricultural University, Hefei 230036)
Abstract:
In order to resolve the problem of image recognition of maize leaf diseases in the field, the Retinex algorithm was used to enhance the image and eliminate the adverse effect of illumination. An automatic threshold method in R-G gray space was adopted for extracting disease spots, the features of color, texture and invariant moments. The principal component analysis method and the support vector machine were used to recognize some common diseases on the maize leaf. The experimental results showed that the total recognition accuracy of southern leaf blight, rust and Curvularia lunata of corn was 90.74%. The study indicated that the proposed method can achieve a good result for disease identification in the natural light condition and it has a certain practical value.
Key words:  maize leaf diseases  uneven illumination  Retinex  principal component analysis  support vector machine

用微信扫一扫

用微信扫一扫