Active-Camera Calibration Using Iterative Image Feature Localization

This work presents a new approach to the problem of calibrating a zooming camera using staged global optimization. The input is a sequence of images of a known calibration target obtained at diier-ent mechanical zoom settings. We avoid heavy dependence on individual, strongly localized features. Feature localization is obtained iteratively, formulated as part of the error criterion used in various passes of the optimization process. The staged optimization process considers all images simultaneously, representing the parameters of the nal calibrated camera as a function of zoom. Our experiments on synthetic images and real stereo image sequences (from a zooming stereo rig) achieve epipolar correspondences at all zoom settings to an accuracy of 1 pixel.