Simulating Visual Pattern Detection and Brightness Perception Based on Implicit Masking
- PDF / 709,306 Bytes
- 11 Pages / 600.03 x 792 pts Page_size
- 66 Downloads / 156 Views
Research Article Simulating Visual Pattern Detection and Brightness Perception Based on Implicit Masking Jian Yang Applied Vision Research and Consulting, 6 Royal Birkdale Court, Penfield, NY 14526, USA Received 4 January 2006; Revised 10 July 2006; Accepted 13 August 2006 Recommended by Maria Concetta Morrone A quantitative model of implicit masking, with a front-end low-pass filter, a retinal local compressive nonlinearity described by a modified Naka-Rushton equation, a cortical representation of the image in the Fourier domain, and a frequency-dependent compressive nonlinearity, was developed to simulate visual image processing. The model algorithm was used to estimate contrast sensitivity functions over 7 mean illuminance levels ranging from 0.0009 to 900 trolands, and fit to the contrast thresholds of 43 spatial patterns in the Modelfest study. The RMS errors between model estimations and experimental data in the literature were about 0.1 log unit. In addition, the same model was used to simulate the effects of simultaneous contrast, assimilation, and crispening. The model results matched the visual percepts qualitatively, showing the value of integrating the three diverse perceptual phenomena under a common theoretical framework. Copyright © 2007 Hindawi Publishing Corporation. All rights reserved.
1.
INTRODUCTION
A human vision model would be attractive and extremely useful if it can simulate visual spatial perception and performance over a broad range of conditions. Vision models often aim at describing pattern detection and discrimination [1– 3] or brightness perception [4, 5], but not both, due to the difficulty of simulating the complex behavior of the human visual system. In an effort to develop a general purpose vision model, the author of this paper proposed a framework of human visual image processing and demonstrated the capability of the model to describe visual performance such as grating detection and brightness perception [6]. This paper will further present a refined version of the visual image processing model and show more examples to investigate the usefulness of this approach. In general, three major issues must be overcome to create a successful vision model. One issue is estimating the capacity of information captured by the visual system, which determines the degree of fine spatial structure that can be utilized by the visual system, which may be modeled by using a low-pass filter. The second issue, the central focus of this paper, is the modeling of nonlinear processes in the visual system, such as light adaptation and frequency masking. It is important to note that the effects of the nonlinear processes are local to each domain. For example, light
adaptation describes the change of visual sensitivity with a background field, the effect of which is limited to a small spatial area [7, 8]. Frequency masking describes the effect of a background grating and occurs, if it does, only when the target and background contain similar frequencies [9]. This space or spatial frequency domain-specific
Data Loading...
