基于色度向量聚类的保险片识别及插接方法  

Recognition and insertion method of automobile insurance tablets based on color vector clustering

作　　者：尤波[1] 陈国杰 梁强 陈潇磊 武坤 You Bo;Chen Guojie;Liang Qiang;Chen Xiaolei;Wu Kun(School of Automation,Harbin University of Science and Technology,Harbin 150080,China)

机构地区：[1]哈尔滨理工大学自动化学院

出　　处：《仪器仪表学报》2018年第7期191-199,共9页Chinese Journal of Scientific Instrument

基　　金：黑龙江省自然科学基金青年基金(QC2014C054)项目资助

摘　　要：针对汽车保险盒的生产所面临的汽车保险片准确识别和快速插接的难题,提出一种基于色度向量聚类的汽车保险片识别及插接方法。该方法使用CCD工业摄像机采集汽车保险片图像信息,通过平均背景法对图像中的背景进行消除后采用污点修复技术进行图像增强,再以不同类别的汽车保险片的色度中心向量为特征提取依据,完成对颜色信息基于色度向量聚类的特征提取,最后通过将提取的特征矢量输入支持向量机,达到对不同类型保险片的识别分类,最终控制SCARA四轴机器人依据分类信息完成精确插接环节。经过大量实验证明,所设计的汽车保险片识别及插接方法对常用的9类汽车保险片识别正确率达到99.7%,在平均插接周期为1 s时的汽车保险盒插接的正确率达95.6%以上。The production of automobile insurance boxes face the challenge on accurately recognizing and rapidly inserting automobile insurance tablets. Aiming to solve the problem,we propose a method identifying and inserting automobile insurance boxes based on color vector clustering. This method uses the industrial CCD camera to obtain image information of automobile insurance tablets. The average background method is utilized to eliminate the image background,and the stain repair technology is used to enhance the image. The center color vector of different kinds of automobile insurance tablets,obtained from image information,are regarded as the basis of features extraction. Then,the feature vector is input into support vector machine（ SVM） so that the classification and identification of different types of insurance can be achieved. Based on the classification results,four-axis robot,SCARA is controlled to complete the accurate insertion. Based on verification experiments,the method of recognizing toward the 9 types of automobile insurance tablets can achieve accuracy rate of 99. 7%. When the average plug cycle is 1 s,the correct rate is above 95. 6%.

关 键 词：平均背景法 污点修复处理 色度向量聚类 支持向量机 SCARA机器人 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术] TH-39[自动化与计算机技术—计算机科学与技术]