自然环境下嫩芽与茶梗自动分割方法的研究  

Study on Method of Automatic Segmentation for Tea Sprouts and Stalk Under Nature Condition

作　　者：姚波[1] 汪洋[1] 李小瑞[1,2] 吕军[1] Yao Bo Wang Yang Li Xiaorui Lv Jun(School of Information Engineering, Huangshan University, Huangshan 245041 College of Information Engineering,Shanghai Maritime University)

机构地区：[1]黄山学院信息工程学院,黄山245041 [2]上海海事大学信息工程学院

出　　处：《黑龙江八一农垦大学学报》2017年第2期114-117,128,共5页journal of heilongjiang bayi agricultural university

基　　金：国家级大学生创新创业训练项目(201410375013)

摘　　要：在茶叶智能采摘过程中,实现嫩芽与鲜梗的自动分割可以减少制茶工序和提高茶叶质量。以自然环境下茶叶嫩芽图像为研究对象,利用G-B灰度图结合直方图阈值法实现了新茶(嫩芽与鲜梗)分割,对新茶二值图像进行形态学腐蚀操作确定嫩芽与鲜梗的分割点,通过逐行扫描实现了自然环境下嫩芽与鲜梗的采摘点标记。实验表明,该方法可以为自然条件下茶叶嫩芽与鲜梗的自动分割提供理论基础。The intelligence segmentation methods of tea sprouts and stalk were studied for reducing the tea-making process and im- proving the quality of tea. Using the images Of tea sprouts under natural condition as the research example, the G-B gray images and histogram threshold method were proposed to achieve automatic segmentation of fresh tea （tea sprouts and fresh stalk）, the morpho- logical erosion operation was used for finding the splitting point between tea sprouts and stalk, the progressive scanning method was presented for labeling the point of intelligent picking. The experiments showed that the method based on image processing could pro- vide a theoretical basis for the automatic segmentation of tea sprouts and stalk under nature conditions.

关 键 词：茶梗 图像分割 腐蚀 

分 类 号：S571.1[农业科学—茶叶生产加工]