改进YOLOv8s的轻量化牛脸识别模型

打开文本图片集

中图分类号：S823；TP391.41 文献标识码：A 文章编号：2095-5553（2025）08-0128-07

Abstract：Inordertorealizenon-contactandreal-time detectionof individualidentityrecognitionofcow faces indairy farmswith complexenvironments，thispaper proposesahigh-performanceand lightweightrecognitionmodelbasedon YOLOv8s target detection network．In this study，17Holsteincowsare takenas theresearchobject，anda video camera isinstalednexttothefeedingchannelofthecowstoobtainthevideoofthecowsatregularintervalsandautomatically， andthefacialimagesof thecowsareobtainedbyvideoframedecompositiontechnology，andthesimilarity betweenthe images ismeasuredbythestructural similarityindex method，sothatthoseimageswithtoohighsimilarityarerejected， andtheindividual numbersof thecowsarethen manualllabeled.In this paper，theYOLOv8smodel isusedas the basis foradding the attntion mechanism CBAM to the backbone network to improve the accuracy of the algorithm，followed by the introductionof the Slim—Neck design paradigm，which replaces the traditional convolutional module（SC）with the GSConv lightweight convolutional module，and replaces the C2f module with the VoV—GSCSP module based on the designofthe GSConvtoreducetheburden onthemodelwhilemaintaining theaccuracy.The model memory footprintof the improved YOLOv8s is 21.3MB ，which is 1.3MB smaller than that of YOLOv8s，and the FPS， P ， R and mAP （20 areimproved by 39.57% ， 5.68% ， 7.74% and 3.33% ，respectively.The improved YOLOv8s can ensure the lightweightandaccuracyof network model，atthesame time has beterrobustness forthefacialrecognitionofcows，and can realize the individual face recognition of cows in dairy farms with complex environments. Keywords：dairy cow breeding；cow face recognition；YOLOv8s algorithm；lightweight;attention mechanism

0 引言

奶牛养殖业是我国畜牧业的重要组成部分，其中奶牛个体身体状况的数字化、精细化、智能化管理已成为现代科学养牛的主要发展方向[1]。（剩余11846字）