自然条件下茶叶嫩芽图像分割方法的研究  

Study on Automatic Segmentation of Tea Sprouts under Natural Conditions

作　　者：张可[1] 吕军[2] 

机构地区：[1]安徽大学计算机科学与技术学院,合肥230601 [2]黄山学院信息工程学院

出　　处：《黑龙江八一农垦大学学报》2016年第2期100-104,共5页journal of heilongjiang bayi agricultural university

摘　　要：探讨茶叶嫩芽自动分割方法,为茶叶智能采摘提供技术支持。以自然环境下茶叶嫩芽图像为研究对象,比较了基于颜色的阈值分割与聚类分割方法对茶叶嫩芽自动分割的影响。首先,选择了R-B和b分量进行茶叶阈值分割;其次在Lab颜色模型下进行K-means聚类分割;最后,通过形态学处理实现茶叶嫩芽自动识别。基于聚类的茶叶分割方法不仅能够抑制颜色阈值分割受光照的影响,且实现了自然环境下茶叶嫩芽的有效分割。Methods of automatic segmentation of tea sprouts were studied in order to provide technical support for the intelligent tea packing. Using the images of tea sprouts under natural conditions as the research examples,the effect of segmentation of tea sprouts based on threshold segmentation and clustering was compared. First,obtaining R-B and b color component to segment image based on threshold met hods. Then,using K-means clustering segmentation method under Lab color space to identify tea sprouts.Finally,detecting the tea sprouts by morphology processing.The results showed that the method based on clustering segmentation could be effective for tea sprouts packing under natural conditions and reduce the influence by light.

关 键 词：茶叶嫩芽 图像处理 自动分割 聚类分析 

分 类 号：S571.1[农业科学—茶叶生产加工]