SELF-CALIBRATION OF STATIONARY CAMERAS

A new practical method is given for the self-calibration of a camera. In this method, at least three images are taken from the same point in space with different orientations of the camera and calibration is co mputed from an analysis of point matches between the images. The metho d requires no knowledge of the orientations of the camera. Calibration is based on the image correspondences only. This method differs funda mentally from previous results by Maybank and Faugeras on self-calibra tion using the epipolar structure of image pairs. In the method of thi s paper, there is no epipolar structure since all images are taken fro m the same point in space, and so Maybank and Faugeras's method does n ot apply. Since the images are all taken from the same point in space, determination of point matches is considerably easier than for images taken with a moving camera, since problems of occlusion or change of aspect or illumination do not occur. A non-iterative calibration algor ithm is given that works with any number of images. An iterative refin ement method that may be used with noisy data is also described. The a lgorithm is implemented and validated on several sets of synthetic and real image data.