3种菟丝子种子数学形态特征的初步研究  

Primary study on mathematic morphological characters of Cuscuta seeds

作　　者：谢敏[1] 沈佐锐[1] 施宗伟[2] 郭琼霞[3] 高灵旺[1] 潘鹏亮[1] 

机构地区：[1]中国农业大学农学与生物技术学院IPMist实验室,北京100193 [2]国家质量监督检验检疫总局,北京100088 [3]福建出入境检验检疫局,福州350001

出　　处：《植物保护》2008年第6期53-56,共4页Plant Protection

基　　金：外来有害生物检疫检测与防范技术体系研究(2006NZ00021);农业公益性行业科研专项(200803022);国家“863”计划(2007AA10Z23701)

摘　　要：本研究基于数字图像处理与分析技术,首次用FeatureExtract数学形态特征提取软件获取了3种菟丝子图像的13项数学形态特征值,并做了方差分析和聚类分析。结果表明,该技术可应用于杂草种子快速鉴定。同时提出对方法的改进建议,为最终实现杂草种子的快速鉴定奠定了基础。Based on the digital image processing and analysis technology, this study firstly applied the geometric character-extracting software to extraction of thirteen parameters of exterior shape feature from three varieties of Cuscuta seeds, which were then subjected to variance and cluster analyses, The results showed that the technology could be applied to rapid identification of weed seeds. This may be a new direction for prompt classification of weed seeds, and provide reliable foundation for the quarantine department to classify weed seeds promptly and accurately.

关 键 词：计算机视觉技术 数学形态特征 检疫性杂草种子 

分 类 号：S453[农业科学—植物保护]