烤烟烟叶自动分级的智能技术  

Intelligent Technologies in Automatic Grading of Flue-Cured Tobacco

作　　者：韩力群[1] 何为[1] 段振刚[1] 苏维均[1] 石芹侠[1] 

机构地区：[1]北京工商大学

出　　处：《农业工程学报》2002年第6期173-175,共3页Transactions of the Chinese Society of Agricultural Engineering

摘　　要：研究烤烟烟叶自动分级中的几种智能技术 ,提出应用图像处理技术提取烟叶质量特征参数 ,应用模糊统计技术确定特征参数对各组与各级的隶属度 ,以及应用综合评判技术判定样本的组别与级别等方法 ,并给出对云南产烟区烟叶样本的分级结果。分级结果表明 ,该系统已达到人类专家分级水平 。Several intelligent technologies used in the automatic grading of flue-cured tobacco were studied. Image processing techniques were used to extract the characteristics of tobacco. Fuzzy pattern recognition statistics were applied to calculate the degree of membership for grading tobacco according to its characteristics, and comprehensive judgement was utilized to determine which group and grade a sample tobacco leaf belongs to. Hundreds of tobacco samples produced in Yunnan Province were graded, which proves that the grading system has met the standards of tobacco experts and possesses intelligence capabilities for accumulating experience.

关 键 词：烤烟 烟叶 自动分级 智能技术 图像处理 模糊模式识别 

分 类 号：S572[农业科学—烟草工业] TP391.4[农业科学—作物学]