浙江农业学报 ›› 2019, Vol. 31 ›› Issue (9): 1516-1522.DOI: 10.3969/j.issn.1004-1524.2019.09.16

• 植物保护 • 上一篇    下一篇

粘虫板害虫自动识别计数研究

包晓敏, 盛家文   

  1. 浙江理工大学 信息学院,浙江 杭州 310016
  • 收稿日期:2019-04-08 出版日期:2019-09-25 发布日期:2019-10-11
  • 作者简介:包晓敏(1965—),女,浙江东阳人,硕士,教授,从事数字农业方面的研究工作。E-mail: 654353162@qq.com
  • 基金资助:
    浙江省重点研发计划(2018C02027)

Research on automatic identification and counting of insect pests on sticky board

BAO Xiaomin, SHENG Jiawen   

  1. School of Information Science and Technology, Zhejiang SCI-Tech University,Hangzhou 310016, China
  • Received:2019-04-08 Online:2019-09-25 Published:2019-10-11

摘要: 针对人工统计黄色粘虫板上害虫数量费时费力的问题,在自适应去除粘虫板背景的基础上,融合颜色空间矩和图像几何形态,完成害虫特征的量化,采用加权决策的方法来识别害虫,构建害虫自动识别计数系统。以米蛾(Corcyra cephalonica)为例,经测试,该系统对黄色粘虫板上米蛾的识别结果与人工统计结果相对误差在7%,能够有效实现对粘虫板上害虫数量的自动统计。

关键词: 粘虫板, 机器视觉, 自动监测, 植物保护

Abstract: In order to solve the arduous and inaccurate problem of manual counting of insect pests in yellow sticky board, based on the adaptive removal of the sticky board background and fusion of color space moments and image geometry, the quantification of pest characteristics were realized, and identification was carried out by weighted decision. Thus, an automatic identification and counting system was constructed. Taking Corcyra cephalonica on yellow sticky board as test materials, it was shown that the relative error between the results of the proposed system and manual counting on sticky board was less than 7%, indicating the feasibility of the proposed system.

Key words: sticky board, machine vision, automatic monitoring, plant protection

中图分类号: