Institute of Transportation Studies at UC Berkeley

Vanishing points of an image contain important information for camera calibration. Various calibration techniques have been introduced using the properties of vanishing points to find intrinsic and extrinsic calibration parameters. This paper revisits the vanishing points geometry and suggests a simple extrinsic parameter estimation algorithm which uses a single rectangle. The comparison with the Camera Calibration Toolbox for Matlab ® shows that the proposed algorithm is highly competitive. The suggested technique is also applied to a realtime pose estimation for an unmanned air vehicle’s navigation in an urban environment. We present a realtime vanishing point extraction algorithm and a pose estimation procedure. The experimental result on a real flight video clip is presented.