资源描述:
《基于边缘检测与Hough变换的车道线检测毕业论文》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、本科毕业论文基于边缘检测与Hough变换的车道线检测LaneLineDetectionbasedonedgedetectionAndHoughtransform3737基于边缘检测与Hough变换的车道线检测摘要随着道路交通的快速发展和汽车保有量的日益剧增,交通安全引起了社会的普遍关注。统计表明,许多的交通事故是由司机疏忽或疲劳驾驶引起的。而汽车辅助驾驶系统正是被设计用来在车辆将要驶离车道或者发生碰撞时警告司机,从而帮助司机安全驾驶,避免交通事故的发生。车道线的检测识别正是实现这一功能的关键技术。本文对车道线检测识别进行了深
2、入的研究。首先,介绍了基于边缘检测和Hough变换的车道线检测这一课题的研究背景和意义,综述了该课题的国内外研究现状,并进行了简要分析。第二,介绍了本课题编程所需要的工具—MATLAB,并简要的介绍了MATLAB的两大特点。第三,介绍图像预处理。该模块包括图像灰度化,图像滤波,边缘检测等相关技术,且对每一种技术所包含的常用方法都做了实验进行对比,并选择出了合适的方法(图像滤波选用中值滤波,边缘检测选用自定义差分算子)。第四,简绍了Hough变换的原理,使用Hough提取并标记车道线。最后,对本文进行了总结,并提出不足之处,总
3、的来说,本文算法能够有效提取车道线,具有较好的鲁棒性。关键词车道线检测MATLAB图像灰度化边缘检测Hough变换37LaneLineDetectionbasedonedgedetectionAndHoughtransformAbstractWiththerapiddevelopmentofroadtrafficandtheincreasinglyexplosioninautoinventory,trafficsafetyhasarousedwidespreadconcern.Statisticsshowthatmostof
4、trafficaccidentscausedbydriversdrowsiness,inattentiveness.Driverassistancesystemsisthedesignusedtobewarningthedriverwhenthevehicleleavingthelaneoracollision,soitcanhelptoavoidtrafficaccidents.Thereal-timelanedetectionandnavigationsystemisalsoanessentialmoduleofthed
5、riverassistanceandwarningsystem.Thispaperisfocusonlanedetectionanddeparturebasedonmachinevision.Firstly,thethesisintroducesthebackgroundandsignificanceaboutthelanedetectionbasedonEdgedetectionandHoughTransform,andthentheresearchstatusofthethesistopicisreviewed.Fina
6、lly,itmakesabriefanalysis.Secondly,thispaperdescribesthetoolsrequiredforprogramming--MATLAB,andmakesabriefintroductiontothetwofeaturesofMATLABThirdly,thethesisdescribethetechnologyrelatedtoimagepre-processingmodulesuchasimagegray,imagefiltering,edgeDetection.Andthe
7、commonmethodsforeachtechnologyincludedhavedoneexperimentstocompareandselectasuitablemethod(Selectingmedianfilteringtofilterimage;Selectingcustomdifferenceoperatorfilteringtomakeedgedetection).Fourthly,makingabriefintroductionofprincipleofHoughtransformandusingHough
8、toextractthelanemarker.Finally,conclusionsaregivenwithrecommendationforfuturework.Overall,theproposedalgorithmcaneffectivelyextractthelane,withbe