基于视觉感知的智能垃圾分拣机器人实验系统设计

打开文本图片集

中图分类号：TN911.73-34；TP242.6 文献标识码：A 文章编号：1004-373X（2025）15-0179-08

Design of intelligent garbage sorting robot experimental system based on visual perception

WU Ling1，²， CAO Qingqing1，²， SHEN Shanshan1，² （1.SchoolofElectronicInformationEngineering/SchoolofIntegrated Circuits，NanjingVocational Universityof Industry Technology，Nanjing，China; 2routitialfr

Abstract：Inthe traditional embedded system courseexperiments，the experimental contentsare singleandunatractie，so itisdifculttostimulatestudents’interestandexplorationdesireefectively.Inviewoftheabove，thispaperdesignsan intellgentgarbagesortingrobotexperimentalsystembasedonvisualperceptionandautonomousnavigation.Inthesystem， RaspberryPiandSTM32aretakenasthecontrolcores，avarietyofsensorsareused toobtain information，andtheMecanum wheelcarisusedas the movingcariertorealizeaseriesoffunctionssuchasautonomousnavigationof therobot，garbage detectionandidentification，colectingandsortingonthebasisof theYOLOv5sobject detectionalgorithm，UWBpositioning solutionndPDmotionfeedbackcontrolalgorithm.Agraphicalhuman-computerinteractioninterfaceisdesigned，andamobile APPcanbeusedtocontroltherobotandobserve itsstatus.The testresultsshowthatthe proposedsystemoperates stablyand reliably，hashighgarbageidentificationaccuracyrate，andhighautonomouspositioningaccuracy.Theexperimentalsystem integratesknowledgeofmultiplecoursessuchasmachinevisionnavigationandpositioning，automaticcontrol，andwireles communication.Ithasgoodopennessandpracticality，andisconducivetocultivatingstudents’engineeingpracticeand innovation capabilities.

Keywords：machine vision;garbage sorting; robot;Raspbery Pi; embedded experimental system;UWB positioning solution

0 引言

传统的嵌人式系统相关课程存在实验内容过时，无法及时反映嵌入式技术的最新发展成果和市场需求；实验设计简单，缺乏挑战性和深度，难以激发学生兴趣和创造力等问题。（剩余8327字）