| 基于阈值分割法的茶叶嫩芽识别研究 |
| |
| 引用本文： | 唐仙,吴雪梅,张富贵,顾金梅.基于阈值分割法的茶叶嫩芽识别研究[J].农业装备技术,2013(6):10-14. |
| |
| 作者姓名： | 唐仙 吴雪梅 张富贵 顾金梅 |
| |
| 作者单位： | 贵州大学机械学院 |
| |
| 基金项目： | 贵州省科学技术基金项目（黔科合J字[2011]2199号）,贵大自青基合2009（035）项目资助 |
| |
| 摘 要： | 嫩芽识别是实现名优茶智能采摘的前提，因此，利用数码相机采集了清明时期贵州大学和贵州农科院茶园茶叶图像，并以茶叶图像为对象，研究了茶叶嫩芽的识别方法。首先对1KGB颜色模型的色差法（R—B）进行分析，然后以该算子为分割分量，对比研究了多种阈值分割法的优缺点和分割效果。试验结果表明，所研究的分割方法都能有效区分嫩芽和背景，其中ostu法和迭代阈值法的识别效果较好，嫩芽识别准确率为90％左右。
|
| 关 键 词： | 茶树图像 茶叶嫩芽识别 色彩空间 阂值法 |
Contrastive research on tender tea recognition based on multiple threshold segmentation methods |
| |
| Authors: | Tang Xian Wu Xuemei Zhang Fugui Gu Jinmei |
| |
| Institution: | (College of Mechanical Engineering, Guizhou University) |
| |
| Abstract: | Identification of the tender tea shoots is the key step towards the intelligent tea harvesting. In this paper, tea images were screened by camera in Guizhou university and Guizhou Academy of tea, with the tea image as the object, identification methods of the tea buds was studied. Firstly tea gray images were analyzed with the color---difference method （R-B） in RGB color spaces. Second, using this method comparative stud- ied the advantages, disadvantages and segmentation results of multiple threshold methods. Finally, effective identification of tender tea shoots can be got. The results show that all methods can distinguish well buds and background, in which ostu and iterative algorithm are best, and the averaged recognition accuracy obtained from using this method is about 90%. |
| |
| Keywords: | tea image tea burgeon identification color space threshold method |
| 本文献已被 维普 等数据库收录！ |
