Geometric Calibration of Digital Cameras. Part I: Camera Model, Subpixel Feature Extraction, and Algorithm Initialization

L. Lucchese (USA)


Camera calibration, pinhole camera, lens distortion, internal and external parameters, perspective projection, pixel aspect ratio.


This paper introduces a new and effective method for deter mining the internal and external geometry of a digital cam era. Traditional camera calibration techniques estimate in trinsic and extrinsic parameters by establishing correspon dences between 3-D target points and their 2-D projections on the image plane of the camera. We show that this can be avoided by mapping features extracted with subpixel ac curacy from various views of a calibration target onto cor responding points of a convenient virtual image that would be generated if the camera were an ideal pinhole devoid of lens distortion and projective warp. The mathematical model for this multi-view alignment technique is presented in this paper along with an algorithm for its initialization; the complete calibration algorithm and its performance are discussed in Part II of this paper.

Important Links:

Go Back