安徽农业大学学报2025，Vol.52Issue(2)：364-372,9.DOI:10.13610/j.cnki.1672-352x.20250519.025
基于机器视觉的单履带自动切割式采茶机
Single-track automatic cutting-type tea picker based on machine vision
摘要
Abstract
[Objective]This study aimed to improve the picking efficiency and quality of tea flushes,a machine vision-based single-track automatic cutting tea picker was designed.[Method]The deep learning model algorithm and image processing were used to achieve the recognition and accurate segmentation of the tea flushes,and the depth point cloud algorithm was integrated to achieve the accurate localization of the pick-ing plane height.The actuator control scheme was designed according to the growth characteristics of the shoots:First,the actuator was moved to the optimum picking plane height based on the position information obtained by the identification system;then,the levelling device and the actuator motion interpolation make the actuator keep moving in the picking plane;finally,the collection of the tea flushes was completed.[Result]The optimal rotation speed of the robotic arm as a test factor for the test,the test showed that the best picking effect was achieved when the rotation speed of the robotic arm was achieved at 37.5 °·s-1.The results of field trials indicated that the complete rate of plucking the tender tips and the rate of the tender tips available for production of the single-track automatic cutting-type tea plucking machine were 82.7%and 85.1%,respectively,which meets the demand for plucking.[Conclusion]This research can enrich the theoretical basis and practical content of mechanized plucking in standardized tea plantations.关键词
深度点云/机器视觉/采摘平面高度定位/路径规划/运动插补Key words
deep point cloud/machine vision/picking plane height positioning/route planning/motion in-terpolation分类
农业工程引用本文复制引用
李为庆,孙燕,曹成茂,陈传涛,张雪晨,赖筱杰..基于机器视觉的单履带自动切割式采茶机[J].安徽农业大学学报,2025,52(2):364-372,9.基金项目
国家重点研发计划项目(2021YFD1601102)资助. （2021YFD1601102）
