基于机器视觉的苹果疏果机器人研究

打开文本图片集

中图分类号：S225；TP242 文献标识码：A 文章编号：2095-2953（2025）05-0043-04

Research on Apple Fruit Thinning Robot Based on Machine Vision

WANG Xiang-yu1， YANG Tie-bin1 * ，WANG Yu-qing²，WANG Hai-bin 1 ， WANG Yang-wei1 （1.Collegeof MechanicalandElectronic Enginering，NortheastForestryUniversity，Harbin Heilongjiang15O040，China; 2.College of Information Technology，Jilin Agricultural University，Changchun Jilin 130118， China）

Abstract：China is a large apple producing country，fruit thinning is time-consuming and labor intensive in the process of apple management，and the research and developmentof fruit thinning robots with automatic fruit thining decision-making can reduce the labor intensityoffruit farmersand improvethe automationlevelof the apple industry.The vision system is the keytechnology to realize intelligent fruit thinning of young apple fruits during the fruit thinning period，and the decision recognition model basedon machine vision is conducive to improving fruithinning decision-making and positioning accuracy.In this paper，YOLOv5n network is used to train the decision-making modeloffruit thinning，obtain thecenter point of young apple fruits and thedecision-making results，combined with Inteal RealsenseD435icamera toobtainthe three-dimensional coordinatevalues ofthe young fruits，and the computer controls the manipulatorto grab the modelofthe young apple fruits according to the coordinate value，which realizes the "hand-eye synergy"between the manipulator and the camera.The experimental results show that the machine vision-based apple young fruit recognition and localization system provides a reliable research basis for the realization of apple fruit thinning robot.

Key words：apple fruit thinning；robot； deep learning；visual recognition；spatial positioning

我国是世界上最大的苹果生产国之一，苹果产业的发展对农村地区的经济增长和农民就业起到了积极的推动作用[1-2]。（剩余4507字）