基于深度学习的观光农业中的桃子采摘识别

打开文本图片集

中图分类号：S24 文献标识码：A 文章编号：2095-5553（2025）07-0153-11

Abstract：Fortheneesofinteligent managementof peach pickingtourismorchards，adeep learning-based picking recognitionmethodis proposed.Themethod usesmachine visionand deep learning technologies tointegratea lightweight humanposture estimationalgorithmLightweight OpenPose，atarget detectionalgorithmYOLOv5s，andatarget tracking algorithm DeepSORT to develop a peach picking behavior detection approach.It can be divided into three steps according tothefunctional order：the picking posture determination method basedonthehuman body joint angles，the pickingtargetdeterminationmethodbasedonnearest neighborretrievaland itsoptimization，andthepicking targetdetection failure solution method based onthesetstatus flags.A dataset is established basedonthe actual peach picking videos for performance tests.Comparing the method based on the angle of human joints proposed in thispaper with the traditionalmethodof using bounding boxesenclosinghuman joints，the methodinthispapercanimprove theprecision of determination rate of hand-raising action by 16% .For the problem of determining the picking target，the nearest neighbor retrieval approach outperforms both thetraditional method basedonthe comparison of distanceand reference size and the method based on the comparison of IoU and thresholds，with an increased P rate by at least 11% . The picking target detection failure solution method basedonsetstatus flags efectivelysolves the influenceof handoclusionon the detection results，substantially improving the P rateby 39% .On this basis，an experimental system is designed to test the proposedmethod under real-world conditions.Theresultsshow that the proposed peach picking recognition method achieves effective and accurate recognition of picking actions in actual orchard environments. Keywords：smart agriculture；agritourism； peach；picking recognition；deep learning； human postures

0 引言

桃子是我国重要的果树品种，也是农业采摘园中常见的果树之一[1]。（剩余18888字）