基于移动目标检测和目标追踪的全局视觉AGV的定位算法  

A Positioning Algorithm for Global Vision Guided AGV Based on Moving Target and Target Tracking

作　　者：刘泽平 刘明兴 麻方达 李晓帆 姚明杰 符朝兴[1] LIU Ze-ping;LIU Ming-xing;MA Fang-da;LI Xiao-fan;YAO Ming-jie;FU Chao-xing(School of Mechanical and Electrical Engineering, Qingdao University, Qingdao 266071, China)

机构地区：[1]青岛大学机电工程学院,青岛266071

出　　处：《青岛大学学报（自然科学版）》2022年第1期85-92,共8页Journal of Qingdao University(Natural Science Edition)

摘　　要：针对目前全局视觉定位算法无法对外形变化较大的自动导引运输车(Automated Guided Vehicle,AGV)进行定位的问题,提出了一种基于移动目标检测和目标追踪的定位方式。选用基于高斯混合模型的背景建模法进行移动AGV检测,当初始选择框太大时采用显著性目标检测方式进一步提高精度,将移动目标检测与目标追踪相结合完成了AGV的定位。实验结果表明,本文设计的定位算法满足实际定位需求,可应用于车间物流AGV。AGV with large shape changes cannot be located by the current global vision positioning algorithm.A positioning method was proposed by moving tcrget detection and target tracking.The background modeling method based on Gaussian mixture model was used to detect the moving AGV.Salient target detection was used to further improve the accuracy when the initial selection box was too large.Then the moving target detection and target tracking was combined to complete the positioning of AGV.Experimental results show that the proposed algorithm can meet the actual positioning requirements and can be applied to AGV of workshop logistics.

关 键 词：AGV 定位 移动目标检测 目标追踪 全局视觉 

分 类 号：TP249[自动化与计算机技术—检测技术与自动化装置]