资源描述：
《Research of Driver Eye Features Detection Algorithm Based on OpenCV》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。
1、2010SecondWRIGlobalCongressonIntelligentSystemsResearchofDriverEyeFeaturesDetectionAlgorithmBasedonOpenCVShifengHu,ZuhuaFang,JieTang,HongbingXuYingSunCollegeofInformation,MechanicalandElectricalCollegeofCivilEngineeringengineeringShanghaiNormalUniversityShanghaiN
2、ormalUniversityShanghai,ChinaShanghai,Chinaz111a222@126.comAbstract—ThispaperproposesadriverfatigueeyefeaturesTheflowchartofdrivereyefeaturedetectionalgorithmdetectionalgorithmbasedonOpenCVimageprocessingandisshowninFig1:computervisiondevelopmentplatform.Thisalgo
3、rithmlocalizeseye-areaanddetectsitsstatebasedonroughtoaccuratethought,andcanlocalizeeyepupilsineye-openstateFigure1.Flowcharofthealgorithmaccurately,whichhassignificancetodecreasetrafficaccidents.Theexperimentshowsthisalgorithmcandetectdrivers’eyestatesaccurately
4、inreal-timeandbeofrobustnessandIII.EYEDETECTIONANDLOCALIZATIONALGORITHMeffectiveness.A.EyeRoughLocalizationKeywordsOpenCV;fatiguedriving;computervision;patternrecognition;Theeyeroughlocalizationinthispaperisbasedontemplatematchingofeyefeaturesafternarrowingthesea
5、rchrange.Itiseffectivetodetecteye-areabyutilizingthisI.INTRODUCTIONalgorithm,whichassuretheaccuracyofeyepupilandstateWiththeincreasingpopularityofautomobiles,fatiguedetection.drivingbecomesanimportantcausetotrafficaccidents,soit1)Pre-processingofeyeroughlocalizat
6、ionismeaningfultoresearchanalgorithmwhichcandetectLocalizingeye-areainawholefaceareawillcausehugedriverfatiguestatesandmakewarningtothetireddrivertocomputation.Ifitcanfurthercutfaceareabeforeeyedecreasethetrafficaccidents.Inthefieldofcurrentfatiguedetectionandnar
7、rowtheeyesearchrangemostly,itwilldrivingdetectiontechnologies,thealgorithmbasedondriversavethecomputationamountandimprovespeedtosatisfyfacefeatureshasbeenoneofthemostpopularmethods.Thistheneedofreal-timerequirement.paperusescomputervision,imageprocessingandpatter
8、nThereisatraditionalproportionruleofgeometricalrecognition,detectseye-areafeaturesafteraccuratefacefeatureoffaces,thatis,thedistancefromhairlinetolocalizationu
