基于YOLOv8的物料视觉检测系统设计OA
Design of a Material Visual Inspection System Based on YOLOv8
在马钢交材炉前智慧物流系统中,输送线托盘上的圆坯数量由物流系统自身提供,缺乏有效的后验证机制.针对此问题,提出一种基于YOLOv8(YOLO是You Only Look Once的简称,v表示版本)的视觉检测系统,用于识别和检测托盘上物料的放置位置及数量,并将识别结果发给智慧物流系统,与物流系统存储的原始数据进行比对,从而确保物料信息的准确性.该视觉识别系统起到了后验证的作用,现场测试结果表明,通过对YOLOv8网络结构进行优化,系统的检测与识别精度超过99.4%,满足了实际应用需求.
In the intelligent logistics system for the furnace front at Masteel Rail Transit Mate-rials Technology Co.,Ltd.,the quantity of round billets on the conveyor line pallets was originally provided solely by the logistics system itself,lacking an effective post-verification mechanism.To ad-dress this issue,a visual inspection system based on YOLOv8(You Only Look Once,v denotes ver-sion)was proposed.This system is designed to identify and detect the placement positions and quan-tities of materials on the pallets.The recognition results are then sent to the intelligent logistics sys-tem for comparison with the original data stored within the system,thereby ensuring the accuracy of material information.This visual recognition system effectively serves as a post-verification step.Field test results demonstrate that by optimizing the YOLOv8 network structure,the system's detec-tion and recognition accuracy exceeds 99.4%,meeting the requirements of practical applications.
朱云峰;耿培涛
宝武集团马钢轨交材料科技股份有限公司,安徽 马鞍山 243000宝武集团马钢轨交材料科技股份有限公司,安徽 马鞍山 243000
信息技术与安全科学
智慧物流YOLOv8实时目标检测圆坯计数深度学习
intelligent logisticsYOLOv8real-time object detectionround billet countingdeep learning
《冶金动力》 2026 (2)
16-21,6
评论