An efficient and practical calibration method for roadside camera using two vanishing points

Abstract

Calibrating roadside camera is essential and indispensable for intelligent traffic surveillance systems. Due to the characteristics of the traffic scenes, the traditional camera calibration methods based on calibration patterns are no longer suitable, since there are generally no calibration patterns (e.g. checkerboard) in traffic scenes. In this paper, we propose a simple and practical calibration method for roadside camera, where the vanishing point in the traffic road direction and the vertical vanishing point are employed that can be easily obtained from most traffic scenes. By making full use of video information, the multiple observations of two vanishing points are available. In order to obtain more accurate calibration results, we present a dynamic calibration method that employs these observations to correct camera parameters and substitutes least squares optimization for closed-form computation. The experimental results on real traffic images demonstrate the effectiveness and practicability of the proposed calibration method.