|
|||||
|
|
| 一种基于支持向量机的烟叶异物检测方法 | |
| 引用本文： | 刘迪清,段圣光,张乐年,沈峘. 一种基于支持向量机的烟叶异物检测方法[J]. 中国制造业信息化, 2012, 41(7): 55-58 |
| 作者姓名： | 刘迪清 段圣光 张乐年 沈峘 |
| 作者单位： | 1. 南京航空航天大学机电学院,江苏南京,210016 2. 山东中烟工业有限责任公司,山东济南,250100 3. 南京航空航天大学能源与动力学院,江苏南京,210016 |
| 摘 要： | 为解决基于颜色空间的检测方法难以剔除与烟叶颜色相同或相近异物的难题,提出了基于支持向量机的检测方法。首先采用Laws算子,对烟叶图像进行预处理,初步分离出目标与背景;然后,根据纹理特征明显与否将待处理区域分为2类;最后,使用支持向量机分别对2个预分类的结果在频域做细分类。试验结果表明,所提方法可以准确地检测出烟叶中的各种异物,误检率较低,满足实际使用要求。 |
| 关 键 词： | 支持向量机 烟叶 Laws 异物检测 |
Detection Method of Foreign Material in Tobacco Based on Support Vector Machine |
|
| LIU Di-qing , DUAN Sheng-guang , ZHANG Le-nian , SHEN Hua. Detection Method of Foreign Material in Tobacco Based on Support Vector Machine[J]. Manufacture Information Engineering of China, 2012, 41(7): 55-58 | |
| Authors: | LIU Di-qing DUAN Sheng-guang ZHANG Le-nian SHEN Hua |
| Affiliation: | g-guang2,ZHANG Le-nian1,SHEN Hua1(1.Nanjing University of Aeronautics and Astronautics,Jiangsu Nanjing,210016,China)(2.China Tobacco Shandong Industrial Co.,Ltd.,Shandong Jinan,250100,China) |
| Abstract: | There are a variety of foreign bodies in tobacco.The existing devices are based on color detection.Some foreign material’s color is familiar with tobacco,and could not be removed.In view of this defect,it proposes a detection algorithm based on support vector machine.This algorithm uses the Laws symbol and filter preprocessing to complete the separation of target and background,and realizes a pre-classification of foreign material.Finally,it takes the support vector machine to classify each of the two pre-classify in the frequency domain.The results show that foreign-based support vector machine high-precision detection method can detect a variety of foreign material,and the false detection rate is low. |
| Keywords: | Support Vector Machine Tobacco Laws Foreign Material Detection |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |