Tire Defect Detection Using Local and Global Features

在线阅读 下载PDF 导出详情
摘要 Inthispaper,wepresentatiredefectdetectionalgorithmbasedonsparserepresentation.Thedictionarylearnedfromreferenceimagescanefficientlyrepresentthetestimage.Astherepresentationcoefficientsofnormalimageshaveaspecificdistribution,thelocalfeaturecanbeestimatebycomparingrepresentationcoefficientdistribution.Meanwhile,acodinglengthisusedtomeasuretheglobalfeaturesofrepresentationcoefficients.Thetiredefectislocatedbyboththeselocalandglobalfeatures.Experimentalresultsdemonstratethattheproposedmethodcanaccuratelydetectandlocatethetiredefects.
作者
机构地区 不详
出版日期 2013年04月14日(中国期刊网平台首次上网日期,不代表论文的发表时间)
  • 相关文献