摘要
茶叶嫩芽识别和采摘点定位是实现精品茶制作过程中机器人选择性自主采摘的前提。提出了一种基于图像处理的自然场景下茶叶嫩芽视觉识别与采摘点定位方法。通过对茶丛图像特征分析,设计了基于超绿特征的茶叶嫩芽图像分割方法,提取茶丛图像的超绿特征,采用大津法(OTSU)进行阈值分割,并通过闭运算去除噪声,经色彩合并获得嫩芽分割图像。设计了结合边缘检测和骨架化处理的嫩芽采摘点定位算法,获得了采摘点在图像中的像素位置,为茶叶嫩芽识别与采摘点定位提供了理论参考。
Recognition of tea flushes and localization of their picking points are the prerequisites of high-quality tea production with the robot to selectively and automatically pick tea flushes.A visual recognition and positioning method based on image processing for tea flushes in natural scenes is proposed.By analyzing on characteristics of tea bush images, a tea flushes image segmentation method based on super-green features is designed.The super-green features of tea bush images are extracted.Threshold segmentation is performed with the OSTU method.The noise is removed by closed operation, and the flushes segmented image is obtained by color combination.Bud picking point positioning algorithm combining edge extraction and skeletonization processing is designed to obtain the pixel position of the picking point in the image, which provide a theoretical reference for the identification and localization of tea flushes.
作者
龙樟
姜倩
王健
朱泓霖
李波
温飞娟
LONG Zhang;JIANG Qian;WANG Jian;ZHU Honglin;LI Bo;WEN Feijuan(School of Engineering,Southwest Petroleum University,Nanchong 637001,China;Nanchong Key Laboratory of Robotics Engineering and Intelligent Manufacturing,Southwest Petroleum University,Nanchong 637001,China)
出处
《传感器与微系统》
CSCD
北大核心
2022年第2期39-41,45,共4页
Transducer and Microsystem Technologies
基金
南充市市校科技战略合作项目(19SXHZ0036,19SXHZ0041)。
关键词
机器视觉
茶叶嫩芽识别
图像处理
图像分割
采摘点定位
machine vision
tea flushes identification
image processing
image segmentation
picking point localization
