首页 | 本学科首页   官方微博 | 高级检索  
     检索      

模拟退火算法在储粮害虫图像识别中的应用
引用本文:张红涛,胡玉霞,邱道尹.模拟退火算法在储粮害虫图像识别中的应用[J].河南农业科学,2003(7):28-31.
作者姓名:张红涛  胡玉霞  邱道尹
作者单位:1. 华北水利水电学院,河南,郑州,450008
2. 郑州大学工学院
基金项目:中国科学院模式识别国家重点实验室开发基金(NLPR2000),河南省自然科学基金(0211030300)
摘    要:简要介绍了储粮害虫智能检测的几个部分:图像采集、图像预处理、特征形成、特征压缩及其分类。对特征选择中的模拟退火算法的思想、实现步骤、参数选择分析等进行了重声、阐述,该算法有效地将储粮害虫的17维原始形态学特征降为10维,提高了分类的效率。

关 键 词:储粮害虫  模拟退火算法  特征选择
修稿时间:2002年11月18

Application of Simulated Annealing Algorithm in Stored-grain Pests Image Recognition
ZHANG Hong-tao,HU Yu-xia,QIU Dao-yin.Application of Simulated Annealing Algorithm in Stored-grain Pests Image Recognition[J].Journal of Henan Agricultural Sciences,2003(7):28-31.
Authors:ZHANG Hong-tao  HU Yu-xia  QIU Dao-yin
Abstract:The stored-grain pests intelligent detection system is introduced briefly, which consists of the pest sample image collecting, image preprocessing, feature forming, feature compressing and the classification. The basic ideal, realization steps, the parameters selecting and others of the simulated annealing algorithm are discussed in detail in features selecting. By this algorithm the 17-dimension morphologic features of the stored - grain pests are compressed into 10 dimensions effectively, thus improves the classification efficiency.
Keywords:Stored-grain pests  Simulated annealing algorithm  Feature selecting
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