| 基于机器视觉的枸杞产地识别研究 |
| |
| 引用本文： | 陈晓峰,张东峰,郝明明,马波,闫修强.基于机器视觉的枸杞产地识别研究[J].农业科技与装备,2013(11):33-35. |
| |
| 作者姓名： | 陈晓峰 张东峰 郝明明 马波 闫修强 |
| |
| 作者单位： | 宁夏大学机械工程学院,银川750021 |
| |
| 基金项目： | 2011年宁夏回族自治区科技攻关项目（411-0114） |
| |
| 摘 要： | 设计以枸杞图像的特征进行产地识别的可行性验证试验。选用3份枸杞样本，在光照箱拍照获得图像，经预处理后得到枸杞形态、颜色及纹理方面的54个特征；运用主分量分析法（PCA）优化得到13个主成分；通过建立支持向量机（SVM）模型，进行样本的训练和测试。结果表明，模型对3个不同产地的构杞正确识别率达到100％。研究结果对于枸杞品种及产地的识别及枸杞谱系研究具有一定参考价值。
|
| 关 键 词： | 机器视觉 枸杞产地识别 主分量分析 支持向量机 |
Study on Origin Recognition Method of Gouqi based on Machine Vision |
| |
| CHEN Xiaofeng,ZHANG Dongfeng,HAO Mingming,MA Bo,YAN Xiuqiang.Study on Origin Recognition Method of Gouqi based on Machine Vision[J].Agricultural Science & Technology and Equipment,2013(11):33-35. |
| |
| Authors: | CHEN Xiaofeng ZHANG Dongfeng HAO Mingming MA Bo YAN Xiuqiang |
| |
| Institution: | (Mechanical Engineering School of Ningxia University, Yinchuan 750021, China) |
| |
| Abstract: | To verify the feasibility of Gouqi origin recognition by Gouqi image characteristics, three Gouqi examples are chosen. After taking photos and preprocessing 54 traits based on morphology, color and texture are collected. There are 13 principal components using PCA. SVM is modeled to train and test Gouqi examples. The results show that SVM model can carry on the distinction of three origins and the accuracy rate of identification is 100%. The results offer valuable reference for Gouqi variety and origin recognition, also for Gouqi pedigree study. |
| |
| Keywords: | machine vision origin recognition of Gouqi PCA SVM |
| 本文献已被 维普 等数据库收录！ |
