object detection using neural networks and genetic programming

《object detection using neural networks and genetic programming》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。

1、VICTORIAUNIVERSITYOFWELLINGTONTeWhareWanangaoteUpokooteIkaaMauiVUWSchoolofMathematics,StatisticsandComputerScienceTeKuraTatauPOBox600Tel:+6444635341WellingtonFax:+6444635045NewZealandInternet:office@mcs.vuw.ac.nzObjectDetectionusingNeuralNetworksandGeneticProgram

2、mingBarretA.ChinSupervisor:Dr.MengjieZhang12October2007SubmittedinpartialfulfilmentoftherequirementsforBachelorofSciencewithHonoursinComputerScience.AbstractObjectdetectionhasbecomeaninterestingandimportantresearchtopicwithincomputerscience.Everydaywearefacinginc

3、reasingvolumesofdatawhereoneofthesetypesofdatathatisofgreatinterestareimages.Thishasledtotheneedforcomputerstosearch,recogniseandclassifyrelevantimagesforallformsofapplicationsfrommilitarytocivil.Thefieldsofneuralnetworksandgeneticpro-gramminghaveachievedthesegoa

4、lswithvaryingsuccess.Thisprojectwillin-vestigatethetaskofeffectiveandefficientobjectdetectionusingneuralnetworksandgeneticprogramming.Thisprojectwillbeginbyexaminingastandardneu-ralnetworksystemandastandardgeneticprogrammingobjectdetectionsystemtodeterminewhichsy

5、stemisfasterandwhatapproachesshouldbetakeninor-dertodevelopamoreeffectiveandmoreefficientobjectdetectionsystem.Intheendthisprojectwillproduceanimprovedapproachthatismoreeffectiveandmoreefficientthanmosttraditionalobjectdetectionsystems.AcknowledgmentsIwouldliketot

6、hankDr.MengjieZhangforhissupport,adviceandtheuseofhisNeuralNetworkpackagetoactasthestandardNeuralNetworksystem.IwouldalsoliketothankWillSmartforprovidingtheuseoftheVGPsystemtoactasthestandardGPsystem.FinallyIwouldalsoliketothanktheGPdiscussiongroupforideasandver

7、yhelpfulsuggestions.iiiContents1Introduction11.1Motivation.......................................11.2ResearchGoals....................................21.3MajorContributions.................................21.4StructureofthisDocument.............................3

8、2Background52.1ComputerVision...................................52.1.1ObjectRecognition..............................52.1.2Objectdetection............................