Calibrating a Linear Visual System by Comparision of Inputs Across Camera/Eye Movements

Abstract

A visual system is calibrated geometrically if its estimates of the spatial properties of a scene are accurate: straight lines are judged straight, angles are correctly estimated, collinear line segments are perceived to fall on a common line. A visual system can fail to be calibrated because of a mismatch between its optics and later visual processing: calibration of computer vision systems typically requires remapping the sensor inputs to compensate for spherical aberration in the camera lens [1].