Research And Implementation Of Lane Departure Warning System

With the increasing auto population, road traffic safety has become a serious problemfor drivers and passengers, and vehicle initiative safety technology has become more andmore significant. Lane Departure Warning System (LDWS) is one of the most popular fieldsof research. It alarms the driver when the vehicle departs from lane center without hisattention, which may cause accidents. Researchers have reported that significant amount ofaccidents can be avoided with the assistant of LDWS. For this reason, LDWS will be aregular equipment in a couple of years. Unfortunately, related technology is underdeveloped.In this context, LDWS is chosen as the subject for this thesis. Works are as follows:1. Propose an edge extraction scheme based on Kalman Filter for lane border detection,which can control the growing direction for it when searching for feature points,unsmooth edges will be removed;2. Propose a voting scheme for vanish line extraction base on the fact that all parallels onthe word plane will intercept at vanish line on the image plane;3. According to the Technical Standard of Highway Engineering and standard for roadtraffic signs and markings, propose a lane borders recognition scheme based on lanewidth and border width using a method of maximum likelihood; This schemeconcentrate on standard lanes, and is suitable for most of the nonstandard lanes as well;4. Carry out lane tracking according to the distribution of lane width and its border width,and adopt an adaptive Kalman Filter for lane width and border width filtering;5. Analyze some basic lane departure warning models, and carry out a lane departurewarning scheme based on the normalized transverse position;6. Test and evaluate the whole system, and conclusion about the performance of the systemis given based on the testing results.