Precise vehicle localization is a basic and critical technique for various ITS applications. It also needs to adapt to the complex road environments in real time. The Global Positioning System (GPS) and the Strap-down Inertial Navigation System (SINS) are two common techniques in the field of vehicle localization. But the localization accuracy, reliability and real-time performance of these two techniques can not satisfy the requirement of some critical ITS applications such as collision avoiding, vision enhancement and automatic parking. Aiming at the problems above, a precise vehicle ego-localization method based on image matching was proposed, which included 4 steps, 1) Calibration. Based on Zhang’s calibration method, the internal and external parameters of the camera were acquired. 2) Image Preprocessing. From the camera parameters, the barrel distortion of the pavement images was corrected, and then the Inverse Perspective Mapping (IPM) operation was executed to the corrected images to get the vertical-view images. 3) Extraction of Feature Points. After preprocessing, the local features in the pavement images were extracted using an improved SURF algorithm. 4) Matching of Feature Points and Trajectory generation. Through the matching and validation of the extracted local feature points, the relative translation and rotation offsets between two consecutive pavement images were calculated, eventually the trajectory of the vehicle was generated. Three scenarios were designed to verify the accuracy of the proposed algorithm. The experimental results show that, the studied algorithm has an accuracy at decimeter-level, and it fully meets the demand of the lane-level positioning in some critical ITS applications.

CICTP