Compensation for Distortion in Eye-Movement Monitors

A compensation scheme for the correction of both horizontal and vertical eye position measurements is developed. Based on the assumption that the distortion process can be modeled as piecewise linear, a new sample point structure is introduced. This structure permits both fixation-based and tracking-based sampling strategies. It also provides a convenient method for partitioning a display for estimation purposes. A minimum mean squared error estimate for the regional linear functions is derived. The problem of measurement assignment to linear regions is solved using a weighted distance metric. Performance is given both in terms of estimation error and the error remaining after correction for a test pattern. Results are presented for a specific example.