The accuracy and reliability of the vehicle positioning system are important performance indices of advanced driver assisted systems and the autonomous driving systems. The paper focuses on the development and verification of vehicle positioning techniques by using Lidar and GPS/IMU sensors. To this end, the techniques for feature extraction, map building, and point cloud matching are investigated. The techniques are then integrated and implemented in a robotic operating system (ROS) platform. Experimental results verify the feasibility of the proposed sensor fusion technique with roadside feature extraction characteristics in rendering high accuracy and reliability vehicle positioning.