Power-spectrum shaping of halftone patterns and its effect on visual appearance

The authors present a binary-pattern power-spectrum-matching algorithm (BIPPSMA) for power spectrum shaping of halftone patterns. This algorithm preserves the first-order statistics of a binary pattern, while manipulating different features of its power spectrum. Binary patterns created with BIPPSMA are presented to human observers during a psychovisual test, in order to measure the response of the human visual system to power-spectrum variations of binary patterns. The characteristics of the power spectrum that are varied in BIPPSMA are the principal frequency location and the energy at the principal frequency peak. Patterns generated with this algorithm and the results of the psychovisual test are shown.<
>