Construction of Virtual Lane and Distinction of Vehicle Position for Driver Assistance System
We present a system that finds road boundaries and constructs the virtual lane based on fusion data from a laser and a monocular sensor, and detects forward vehicle position even in no lane markers or bad environmental conditions. When the road environment is dark or a lot of vehicles are parked on the both sides of the road, it is difficult to detect lane and road boundary. For this reason we use fusion of laser and vision sensor to extract road boundary to acquire three dimensional data. We use parabolic road model to calculate road boundaries which is based on vehicle and sensors state parameters and construct virtual lane. Due to the distinct features of two types of vehicle (opposite direction and same direction), different approaches are necessary to detect each direction. And then we distinguish vehicle position in each lane.
Vehicle detection adaboost 3d optical flow haar-like road boundary detection
Hyun-Koo Kim Muzaffar Djalalov Nguyen Khac Cuong Ho-Youl Jung Ju H. Park
国际会议
The International Conference on Electrical Engineering 2009(2009 电机工程国际会议)
沈阳
英文
1-6
2009-07-05（万方平台首次上网日期，不代表论文的发表时间）