On Zoom-Dependent Calibration of Cosumer-Grade Cameras

The objective of geometric camera calibration is to determine a parametric model of the projective features of an imaging device. These features describe the mapping of the image coordinates to the viewing rays in space and are of importance for various applications in computer vision like synthetic views, spatial reconstruction and tracking. Numerous work exist on solving the problem for fixed focus lenses. In this book especially the problem of calibrating cameras with zoom optics is addressed, since this a problem rarely dealt with in literature. What makes it so challenging is the high dependency of various camera model parameters on zoom levels and their determination throughout a continuous range of zooming and tracking. In the first two sections of this work the fundamentals of projective geometry and applied optics, neccessary to understand our approach, are explained. Following parts introduce a feasable method for solving the problem and present our experiments conducted with consumer-grade SLR cameras.

Autorentext

Helmut Zollner, born in Austria (1973), studied Computer Science at Vienna University of Technology, specialises in Computer Vision and Pattern Recogintion. (M.S.) Currently the author works as algorithm engineer for camera-based driving assistance systems at an international automotive company in Germany.