欢迎来到天天文库
浏览记录
ID:33390341
大小:3.35 MB
页数:66页
时间:2019-02-25
《智能视频监控中的几个关键问题》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库。
1、AbstractInrecentyears,theintelligentvideosurveillancehasbeenahotspotofcomputervisionresearch.Theintelligentvideosurveillancetechnologyhasbeenwidelyappliedtohandlevariousspecificproblems.Thispaperfocusesonseveralkeyissuessuchastrip-linedetection,abandoneddetectio
2、nandpicturerectificationforclassroomvideosurveillance.Relevanttheoriesandalgorithmstosolvetheproblemsintheseissuesarepresentedinthispaperindetailsandtheintroductionofourtrip-linedetectionsystemisalsogiven.Thetrip-linedetectiontechniquediscussedinthispapercanbeap
3、pliedforthesecurityprecautionsofcommunitiesorschools.Onthisissue,manymovingobjectdetectionalgorithmsarestudied,includingGaussianmixturemodeling,BayesianmodelingandViBemodeling.Theremoveofshadowbasedoncolorinformationandeliminationofthefalsedetectioncausedbysudde
4、n-lightchangesareinvestigatedinthispaper.Furthermore,thebasictheoryofKalmanfilterandCamShiftforobjectstrackingisstatedinthispaper,andthenboththesetwomethodsarecombinedtosolvemulti-objectproblem.Finally,amethodfordeterminingthetrip-lineanditsdirectionisgiveninthi
5、spaper.Whilethekeypointforabandonedobjectdetectionliesindistinguishingthemovingforegroundfromthestaticforeground,multi-levelforegroundsegmentationisusedtosolveitinthispaper.Ontheotherhanditiscriticaltodifferentiatetheabandonedobjectsfromthestaticforegroundwithou
6、tthebackgroundremovedones.Thecontoursimilaritycombinedwiththecolorhistogramsimilarityisadoptedinthispaper.Thenstatichumanbodiesfromthestaticobjectshavetobeexcluded,andSVM+HOGmethodisadoptedtodealwithit.Therectificationofthedistortedclassroomroommonitorimagecanbe
7、appliedforexaminationsupervision,andthegoalofthisrectificationistoremovethelensdistortionandperspectivedistortioncausedbythepositionofseats.Thisthesisdescribesarectificationmethodforcameracalibrationandalsodescribesacretinmeanstorectifyperspectivedistortionofthe
8、picturesinthisapplication.Keywords:IntelligentVideoSurveillanceTrip-lineDetectionAbandonedDetectionDistortedPictureRectification目录第一章绪论...............................
此文档下载收益归作者所有