安徽农业科学2025，Vol.53Issue(8)：200-204,208,6.DOI:10.3969/j.issn.0517-6611.2025.08.041
YOLOv8算法分类识别机收甘蔗杂质的研究
Research on YOLOv8 Algorithm Classification and Recognition of Sugarcane Impurities in Machine Harvesting
摘要
Abstract
In view of the sugar factories rely entirely on manual judgment of the impurity content in mechanically harvested sugarcane,which is highly subjective and lacks scientific basis,used deep learning methods to explore the classification and recognition of impurities in mechani-cally harvested sugarcane.The main focus had been build an image acquisition platform and dataset,selected,laid out,and designed hard-ware such as computers,industrial cameras,and light sources.The YOLOv8 algorithm had been used to train and detect the dataset for classi-fication,used recall,precision,and average precision mean mAP@0.5 quantitative evaluation of the detection results,showed that the YOLOv8 algorithm achieved an average accuracy of 77.4%in classifying and recognizing machine harvested sugarcane,effectively distinguis-hing its different components.This technology research will lay a certain foundation for the subsequent detection of impurity content in machine harvested sugarcane.关键词
机收甘蔗/YOLOv8/杂质/分类/识别Key words
Machine-harvested sugarcane/YOLOv8/Impurity/Classify/Identify分类
农业科技引用本文复制引用
周思理,李国杰,何冯光,郑爽,代叶,邓干然..YOLOv8算法分类识别机收甘蔗杂质的研究[J].安徽农业科学,2025,53(8):200-204,208,6.基金项目
湛江市科技计划项目(2021A05188) （2021A05188）
海南省自然科学基金项目(522QN385) （522QN385）
海南省自然科学基金项目(324MS095) （324MS095）
广西壮族自治区市场监督管理局科技项目(GSJKJZC2024-3). （GSJKJZC2024-3）
