视觉机器人货物识别技术探究  

Research on Visual Robot Cargo Identification Technology

作　　者：田建涛[1] 苏沛[1] 王娜[1] 

机构地区：[1]长安大学道路施工技术与装备教育部重点实验室,西安710064

出　　处：《北京联合大学学报》2014年第4期12-17,共6页Journal of Beijing Union University

摘　　要：对某现代物流储仓中168箱货物进行机器视觉技术试验,通过硬件和软件系统对货物进行了图像采集、处理和分析,建立机器人视觉技术进行货物物流公司和包装类型的识别模型,并对该机器人视觉识别模型进行检验,综合识别准确率高达98.96%。该视觉系统可有效用于现代物流储仓的机器人货物识别,能够加快物资转运速度、降低运作成本并使控制自动化,符合现代物流的要求,市场前景广阔。168 cartons in a modern logistics warehouse has been tested with the machine vision technology. Image acquisition, processing and analysis of the goods has been done with hardware and software system, and recognition model of robot vision technology for cargo logistics company and packing type has been established. Test shows that the robot vision recognition model makes the comprehensive recognition accuracy rate as high as 98.96%. The vision system can be effectively used for robot goods identification of modern logistics warehouse, and can accelerate the material transfer speed, reduce the operation cost and the control automation, and meet the requirements of modern logistics as well. Its market prospect is broad.

关 键 词：视觉技术 货物识别 图像处理 物流储仓 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]