Abstract

A line-element target differing sufficiently in orientation from a background of line elements can be visually detected easily and quickly; orientation thresholds for such detection are lowest when the background elements are all vertical or all horizontal. A simple quantitative model of this performance was constructed from three processing stages: (1) linear filtering by two classes of anisotropic filters, (2) nonlinear point transformation, and (3) estimation of a signal-to-noise ratio based on responses to images with and without a target. A Monte Carlo optimization procedure (simulated annealing) was used to determine the model parameter values required for providing an accurate description of psychophysical data on orientation increment thresholds.

© 1995 Optical Society of America

Spatial alignment across gaps: contributions of orientation and spatial scale

Sarah J. Waugh and Dennis M. Levi
J. Opt. Soc. Am. A 12(10) 2305-2317 (1995)

Optimum distortion-invariant filter for detecting a noisy distorted target in nonoverlapping background noise

Bahram Javidi and Jun Wang
J. Opt. Soc. Am. A 12(12) 2604-2614 (1995)

Two-dimensional approach to psychophysical orientation tuning

James L. Dannemiller and James N. Ver Hoeve
J. Opt. Soc. Am. A 7(1) 141-151 (1990)

Differences between stereopsis with isoluminant and isochromatic stimuli

David R. Simmons and Frederick A. A. Kingdom
J. Opt. Soc. Am. A 12(10) 2094-2104 (1995)

Spatial variability as a limiting factor in texture-discrimination tasks: implications for performance asymmetries

Barton S. Rubenstein and Dov Sagi
J. Opt. Soc. Am. A 7(9) 1632-1643 (1990)