基于图像的驾驶主动安全前防撞的方法研究与实现-计算机应用技术专业论文.docxVIP

优秀毕业论文 精品参考文献资料 研究生优秀毕业论文 硕士学位论文 硕士学位论文 基于图像的驾驶主动安全前防撞的方法研究与实现 Abstract C011ision aVoidance of preceding Ve}1icles is an impo咖t research dlrectlon m the fleld of active saf-ety driving，t11e core problem is the detection，recognition and tracking of fonVard vehicles．AJld the main processing operation of detection，recognition and tracl【ing of preceding vehicles is perfomed for the rear fke of Vehicles，this paper mainly focuses on the detection，recognition a11d tracl(ing of the rear f．ace of Vehicles，the main contents are as follows： (1)A coarse positioning method of the rear f．ace of vehicles baSed on the shadow undemeath a vehicle is presented．In tIlis paper，a modifled Gaussian mixture model clustering algorithm is used to cluster targets in traffic images such aS the roads，lane lines，Vellicles aIld shadow undemeath a vehicle，shadow thresh01d is calculated according to the mean and varance of the Gaussian shadow model，thus both the shadow image and the intersecting lines of the vehicle shadow aIld the road are generated，flrLally the initial position of the rear face of vehicles is located by me曜ing t11e inf0硼ation of the aspect ratio of t11e Vehicle aIld the shadow lines．Tllis method can not only efrectively locate me rear f．ace of a Vellicle but also adapt to the change of the intensi够of light at di恐rent times d嘶ng the day and the interference of me extemal enviro衄ent． (2)T11is paper studies two different rear f色a“lre extracion methods of a car of Haar-like and HOG，proposes an improved vehicle verification method based on Haanlike and Adaboost，this metllod can efI’ectively eliminate false alarms generated in the coarse positioning stage．This aIticle uses Adaboost and SVM classifier analyzes me di疵rences and the vellicIe identification result of me two feature ex仃action methods．In order to better hjghIight the horizontal and Vertical edge features of the rear face of Vellicles，