基于机器视觉苗木自动分选的研究

打开文本图片集

中图分类号：TP242；S776 文献标识码：A 文章编号：2095-2953（2025）03-0056-04

Research on Automatic Seedling Sorting Based on Machine Vision

ZHANG Wei-guo 1，2，3 ， LI Shi-fang 1，2，3 ， ZHANG Bei-hang 1，2，3 ， MA Ling 1，2，3 ， YI Hong-lei 1，2，3 *

（1.HarbinReseachIsiuteofFrestryacinerytateFrestrydrassnddmstratioHarbinHeloingOia;2.ForestryElecticalndchnicalEineringab，tateorstrGssanistratioainloiga;3.Forestry Technical Equipment Engineeringand TechnologyResearch Center，State Forestryand GrasslandAdministration， Harbin Heilongjiang ，China）

Abstract：Seedling sorting is an important link of an automatic production line for factory -- style seedling cultivation withtheaim to eliminate theseedlings thatare notsuitablefor cultivationontheautomatic production line through theautomatic sorting of seedlings.Automatic photography，analysis，and judgment of seedlings are carried out byusing machine vision technology，andunsuitable seedlings are sortedout，while the coordinate parameters of these seedlingsare calculated.Throughacommunication protocol，the parameters aresent to thecontrol terminal of the robotic arm，enabling the roboticarm to perform accurate picking actions，pick out the unsuitable sedlings，and place thematthedesignated location，thuscompleting the sorting processAchieving this purpose mainly involves three key technologies，that is，machine vision recognition technology，stacking of the robotic manipulator and communication control.Afterthe completion of theexperimental prototype and the implementation of seedling sorting tests，accurate sorting can beachievedaccording to the setrequirements.Whenapplied to seeding production，ithas practical significance for improving production efficiency and solving the difficulties in hiring workers.

Keywords：machine vision；automatic sorting；stacking

机器视觉是一种利用计算机视觉和机器学习的技术，使计算机系统能够识别影像数据，包括了从数字图像或视频中提取信息、分析和理解图像内容的方法。（剩余4060字）