约束聚类算法的多目标QR码定位识别方法  

METHOD OF MULTI-QR CODE LOCALIZATION BASED ON CONSTRAINED CLUSTERING ALGORITHM

作　　者：王华[1,2] 赵伟甫 张爽[2] 高金刚[2] 刘孝峰 Wang Hua;Zhao Weifu;Zhang Shuang;Gao Jingang;Liu Xiaofeng(School of Mechanical and Electrical Engineering,Changchun University of Technology,Changchun 130012,Jilin,China;School of Mechanical and Electrical Engineering,Changchun Institute of Technology,Changchun 130012,Jilin,China;Ministry of Industrial Planning and Development,CRRC Changchun Railway Vehicles Co.,Ltd.,Changchun 130062,Jilin,China)

机构地区：[1]长春工业大学机电工程学院,吉林长春130012 [2]长春工程学院机电工程学院,吉林长春130012 [3]中车长春轨道客车股份有限公司工业规划发展部,吉林长春130062

出　　处：《计算机应用与软件》2023年第5期243-247,271,共6页Computer Applications and Software

基　　金：吉林省科技发展计划项目(20200401107GX,20190302024GX);吉林省教育厅“十三五”科研规划项目(JJKH20200611KJ)。

摘　　要：在轮对智能选配系统中,为提高轮对零部件的录入效率,提出多目标环境下对单个QR码的定位识别算法。针对使用聚类算法对其进行具有二维码特征的最优聚类问题,通过预先设定簇个数以及初始聚类点,并在其基础上加入单个QR码的特征作为约束,对聚类算法进行改进。实验表明,相比于现有技术方案,在多目标QR码识别需求下,此方法能够降低QR码识别产品的生产成本,识别率达96%,实现了多目标QR码的同时识别。In the intelligent matching system of wheelset,for improving the input rate of wheelset parts,a localization recognition algorithm for a single QR code is proposed in multi-QR environment.Aimed at the problem of using clustering algorithm to get optimal clustering with 2D code features,the clustering algorithm was improved by setting the amount of clusters and the initial clustering points and adding the features of a single QR code as constraints.The experiments show that compared with existing technology,this method can reduce the production cost under the requirement of multi-QR code recognition,and the recognition rate reaches 96%,which realized the simultaneous recognition of multi-QR code.

关 键 词：多目标QR码 约束聚类算法 直线搜索 分割 识别 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]