首页 | 本学科首页   官方微博 | 高级检索  
     

基于颜色和形状特征的黄瓜霜霉病自动识别研究
引用本文:王雪,马卓,王欣,姜丹,刘小溪,贾志雷. 基于颜色和形状特征的黄瓜霜霉病自动识别研究[J]. 安徽农业大学学报, 2013, 40(6): 1071
作者姓名:王雪  马卓  王欣  姜丹  刘小溪  贾志雷
作者单位:吉林省科学技术信息研究所,长春 130021,吉林省科学技术信息研究所,长春 130021,吉林省科学技术信息研究所,长春 130021,吉林省科学技术信息研究所,长春 130021,吉林省科学技术信息研究所,长春 130021,吉林省科学技术信息研究所,长春 130021
基金项目:吉林省世行贷款农产品质量安全项目(2011-Z20)资助。
摘    要:利用计算机图像处理技术,对农作物病害的自动识别做了实验性研究。以黄瓜霜霉病为例,采集了黄瓜霜霉病的数字图像作为试验样本,进行了灰度化处理、二值化处理、平滑化处理和边缘检测处理等,从病斑的形状和颜色两方面提取特征参数,利用k近邻法对样本进行分类识别,识别率达到95%。

关 键 词:计算机图像处理; 霜霉病;k近邻法; 模式识别
收稿时间:2013-04-11

Automatic recognition of cucumber downy mildew based on color and shape feature
WANG Xue,MA Zhuo,WANG Xin,JIANG Dan,LIU Xiao-xi and JIA Zhi-lei. Automatic recognition of cucumber downy mildew based on color and shape feature[J]. Journal of Anhui Agricultural University, 2013, 40(6): 1071
Authors:WANG Xue  MA Zhuo  WANG Xin  JIANG Dan  LIU Xiao-xi  JIA Zhi-lei
Affiliation:Institute of Scientific and Technical Information of Jilin, Changchun 130021,Institute of Scientific and Technical Information of Jilin, Changchun 130021,Institute of Scientific and Technical Information of Jilin, Changchun 130021,Institute of Scientific and Technical Information of Jilin, Changchun 130021,Institute of Scientific and Technical Information of Jilin, Changchun 130021 and Institute of Scientific and Technical Information of Jilin, Changchun 130021
Abstract:In this paper, automatic identification of cucumber downy mildew was conducted using computer image processing. The digital image of the cucumber downy-mildew was collected, and then treated with gray processing, binary processing, smoothing processing and edge detection processing. Characteristic parameters were extracted from the two aspects of shape and color of the lesion, and then k-nearest neighbor method was used for classification and identification. The recognition rate was up to 95%.
Keywords:computer image processing   downy mildew   k-nearest neighbor method   pattern recognition
点击此处可从《安徽农业大学学报》浏览原始摘要信息
点击此处可从《安徽农业大学学报》下载全文
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