This paper introduces a new method to recover 3-D road plane from its 2-D monocular perspective image. The research is aimed at the reconstruction of depth information from the 2-D visual input in road following and navigation. Planar road model is considered and the road-centered coordinate system which forms slope and turn angles with camera-centered coordinate system is used to describe boundary points on road plane. We develop approaches to find matching points of boundaries of road and to obtain angular parameters thereafter. A way of finding depth of matching points from the perspective images and angular parameters together is proposed. Therefore the 3-D road reconstruction can be replicated without introducing any parameters of inverse perspective.

Correspondence problem in road image sequence is discussed and a method to establish road correspondence from its perspective image sequence is suggested. The proposed method is mainly based on the features of turn angles of road edge points, while the turn angle for each edge point at one time can be computed from the frame based on the determination of matching points whin that frame. The turn angles will change from frame to frame according to the panning rotation of the camera and, each stationary edge point, the difference of turn angles between two frames equals the panning angle of the camera. Thus we develop an algorithm to estimate the value of panning angle of the camera by which correspondence in road image sequence can be established.