摘要
以设计茶树害虫成虫智能识别应用系统为目的,提出了构建基于图像特征组合智能识别关键技术的三维空间结构(特征组合可分性结构、算法切合度和识别率)研究思想,系统地阐述了三维空间结构的内涵、指标体系、筛选条件、研究模型及实践方法;以尺蛾科典型害虫成虫为实例,认证了研究办法和实现途径的可行性,并对构建智能识别关键技术之三维空间结构知识库的实现路线图进行了展望。
To design application system of intelligent identification of the typical geometridae adults in tea pests , this paper presented a kind of research idea of three-dimension special structure of intelligent identification . T he key technology is based on image features such as separability structures ,algorithm fit and identification rate .It was systematically elaborated on aspects of the implication ,indicators system ,screening conditions ,researching model and the practices of this three-dimension special structure . Identifying the typical moth in family geometridae of tea pests was cited as a feasibility study example of research method and implementation way . In addition , the realization of intelligent identification key technology knowledge was previewed in this paper .
出处
《浙江科技学院学报》
CAS
2014年第3期192-198,共7页
Journal of Zhejiang University of Science and Technology
基金
国家科技基础性工作专项(2008FY210500)
浙江省教育厅科研计划项目(Y201224733)
关键词
茶树
尺蛾科害虫
图像特征
三维空间结构
智能识别
tea tree
geometridae
image features
three-dimension special structure
intelligent identification