首页 | 本学科首页   官方微博 | 高级检索  
     

农田害虫色彩特征提取的研究
引用本文:邱道尹,张法全,张红涛.农田害虫色彩特征提取的研究[J].华北水利水电学院学报,2004,25(1):30-32.
作者姓名:邱道尹  张法全  张红涛
作者单位:华北水利水电学院,动力系,河南,郑州,450011;郑州轻工业学院,信控系,河南,郑州,450002
基金项目:中科院模式识别国家重点实验室开放基金项目(NLPR2000)
摘    要:为及时准确地监测预报农田害虫的发生情况,采用了图像处理的方法对诱集到的害虫进行识别.选用HSI彩色模型空间,确定了红、黄、绿、青、兰、紫、黑7种主要颜色,每种颜色分别提取了5种形态特征,包括新提出的描述颜色的一个重要特征———圆度的新定义.采用VC++6 0语言编制了程序,对绿刺蛾和红缘灯蛾进行了色彩特征的提取,识别率达到85 7%.结果表明,所提取的特征有利于识别害虫.

关 键 词:彩色模型  RGB空间  HSI空间  特征提取
文章编号:1002-5634(2004)01-0030-03
修稿时间:2003年6月1日

Study on Extracting Agricultural Pests' Color Features
QIU Dao_yin,ZHANG Fa_quan,ZHANG Hong_tao.Study on Extracting Agricultural Pests' Color Features[J].Journal of North China Institute of Water Conservancy and Hydroelectric Power,2004,25(1):30-32.
Authors:QIU Dao_yin  ZHANG Fa_quan  ZHANG Hong_tao
Affiliation:QIU Dao_yin~1,ZHANG Fa_quan~2,ZHANG Hong_tao~1
Abstract:Pests lured through image process are identified in order to monitor and forecast agricultural pests' situation in time. Selecting HSI color model space, seven main colors are determined,such as red, yellow, green, cyan, blue, purple and black. To each color five morphologic features such as area, perimeter, width, height, roundness are extracted,and a new definition of roundness as a main feature on describing colors is put forward. By programming with VC++ 6.0, color features to Parasa consocia and Amsacta lactinea are extracted. The rate of recognition is up to 85.7%. Results show that extracted color features are advantageous to recognizing agricultural pests.
Keywords:color model  RGB  space  HSI  space  feature extraction
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