Testing Simulation of Foveal Vision
A Comparison of Image-Quality Models and Metrics Predicting Object Detection
Visual Effects of the Black Matrix in Tessellated Displays
The Visibility of Higher-Level Jags
Imagery Interpretability Rating Scales
Continuous Assessment of Image Quality
Testing the Value of Gray-Scaling for Images of Handwriting
Virtual Printer: A Software Testbed for Printer Simulation
CRT-Display Simulation of Printed Output
Metrics for Evaluating Printed Map Quality
The Relationships Governing Image Quality in Some Text and Photographic Images
Visual Suppression of One Eye's View with Dichoptic Stimulation
3-D Stereo Displays: How to Move in the Third Dimension
Depth and Size Judgments within Uncollimated and Collimated Stereoscopic Displays
Time-to-Contact Estimation in a Driving Simulator
Visual Requirements for Virtual-Environment Generation
Binocular Vision in Stereoscopic Virtual-Reality Systems
Perspective Projection and Its Tolerances in Partial- Overlap Binocular Head-Mounted Displays
Statistical Study of Eye Movement When Using a Head-Mounted Display
Motion Processing of Stereoscopic Depth Information
Visual Characteristics of Cell-Gap Spacers in a PDLC Projector
Color Nameability Surface for Computer-Displayed Colors. I: Physiological Mechanisms
Testing Simulation of Foveal Vision
E. Peli
Schepens Eye Research Institute, Harvard Medical
School, Boston, MA full paper
Simulations of central vision based on individual CSF functions can be used to predict the maximal viewing distance for which two images will be discriminable. The simulation method, together with a testing paradigm, can provide us with the necessary tools to determine the image-dependent parameters that affect discrimination performance.
A Comparison of Image-Quality Models
and Metrics Predicting Object Detection
A. M. Rohaly
U.S.
Army Research Laboratory, Aberdeen Proving Ground, MD
A. J. Ahumada,
Jr., A. B. Watson
NASA/Ames Research Center, Moffett Field, CA full paper
Many models and metrics for image quality predict image discriminability, the visibility of the difference between a pair of images. When three such methods were compared for their ability to predict the detectability of objects in natural backgrounds, it was found that the channel models performed better than simpler models.
The three methods examined include: a Cortex transform model with within-channel masking, a Contrast Sensitivity filter model, and digital image difference metrics. Each method was implemented with three different summation rules: the root mean square difference, Minkowski summation with a power of 4, and maximum difference. The Cortex model with a summation exponent of 4 performed best.
Visual Effects of the Black Matrix in
Tessellated Displays
J. Larimer
NASA/Ames Research
Center, Moffett Field, CA
J. Gille
Western Aerospace
Laboratories, Moffett Field, CA
R. Martin
Xerox PARC, Palo Alto,
CA
J. Lubin
David Sarnoff Research Center, Princeton, NJ full paper
The ViDEOS-Sarnoff model of human vision was used to analyze the visibility and effects of the black matrix in tessellated displays. Computational experiments relate the visibility of the black matrix to spatial and gray-scale quantization, logical pixel geometry, and aperature ratio. Results are compared to empirical measurements.
The Visibility of Higher-Level
Jags
A. C. Naiman
Hong Kong University of Science
and Technology, Kowloon, Hong Kong
W. Makous
University of
Rochester, Rochester, NY full paper
We have extended our preliminary investigation of visual sensitivity to jagged edges. Agreement amongst data from three subjects allows us to predict sensitivity and leads to a model based on jag levels. We find that only first- and second-level jags play a significant role in the visibility of jaggedness.
Imagery Interpretability Rating Scales
L. A. Maver, C.
D. Erdman, K. Riehl
Itek Optical Systems, Lexington, MA full paper
An imagery interpretability scale is a tool used by people to make and communicate quantitative judgments about the potential interpretability of an image. The aerial imaging community utilizes the National Imagery Interpretability Rating Scale (NIIRS) to define and measure the quality of images and performance of imaging systems. Through a process referred to as "rating" an image, the NIIRS is used by imagery analysts to assign a number which indicates the interpretability of a given image.
The NIIRS concept provides a means to directly relate the quality of an image to the interpretation tasks for which it may be used. Although the NIIRS has been primarily applied in the evaluation of aerial imagery, it provides a systematic approach to measuring the quality of photographic or digital imagery, the performance of image capture devices, and the effects of image processing algorithms. This paper describes the NIIRS and provides information on how the scale is developed and used to facilitate its application to other imaging scenarios.
Continuous Assessment of Image Quality
R. Hamberg, H.
de Ridder
Institute for Perception Research, Eindhoven, The
Netherlands full paper
A new method for assessing time-varying quality is presented. Subjects continuously indicate the perceptual image quality by moving a slider along a graphical scale. The usability of this method is illustrated by an experiment in which subjects assessed the quality of still pictures comprising time-varying degrees of sharpness.
Testing the Value of Gray-Scaling for Images of Handwriting
J. Bedford-Roberts, D. Reynolds
Hewlett-Packard Laboratories,
Bristol, U.K. full paper
Experiments were carried out to test greyscaling as a way to improve the perceived quality and legibility of hand written input on low resolution liquid crystal displays. Results show that greyscaling is equivalent to a significant improvement in screen resolution and indicate the potential for further improvement by small (sub-pixel) realignments of the original images.
Virtual
Printer: A Software Testbed for Printer Simulation
D.
Kalra
Hewlett-Packard Laboratories, Palo Alto, CA full paper
A comprehensive testbed for the simulation of a number of parameters affecting print quality, such as the characteristics of ink, paper, and their interaction, as well as the motion of the printer head and the motion of paper through the printer, will be described. The output of the simulation may be obtained either as numerical output, printed on a high-resolution printer, or displayed on a CRT.
CRT-Display
Simulation of Printed Output
E. R. Anthony and J. E.
Farrell
Hewlett-Packard Laboratories, Palo Alto, CA full paper
An apparatus capable of displaying images with 1200 dpi resolution 24 bit color is described. This display was used to simulate images printed at 200 dpi with varying levels of gray. Observers ratings of image quality for printed images are compared with their ratings of displayed simulations. For both printed photographic images and displayed simulations, optimal image quality was obtained when graylevels were equally spaced in L*, and subjective image quality increased with increasing graylevels up to 32 levels beyond which observers could detect no perceptible change. The results demonstrate that it is possible to simulate printed output on a 1200 dpi, 24 bit color display.
Metrics for
Evaluating Printed Map Quality
H. Veron, J. R. Leger
The
MITRE Corp., Bedford, MA full paper
A set of checkerboard test patterns embedded in a digital map is used to rapidly assess the many components of a hardcopy print system. Spatial frequency analysis of the patterns provides the metrics that correlate with observer assessments of map quality. This work will demonstrate how the test patterns and metrics are used to evaluate compression schemes, filters, printers and printer papers.
The Relationships Governing
Image Quality in Some Text and Photographic Images
N. W.
Burningham
Eastman Kodak Co., Rochester, NY
full paper
Fundamentally, the quality of reflection hard copy is governed by visually significant attributes of the image such as tone scale, noise, sharpness, and uniformity. Writing resolution, addressability, and bit depth of gray-level imaging are insufficient specifications of quality, taken individually. Studies of the image quality of text and of color pictorial images, digitally produced with varying resolution and number of gray levels, illustrate the interactions of these parameters in producing varying levels of perceived quality.
Visual
Suppression of One Eye's View with Dichoptic Stimulation
R. Blake
Vanderbilt University, Nashville, TN full paper
When the two eyes receive different stimulation (which can occur with artificial viewing devices), binocular single vision gives way to alternating periods of monocular suppression; this outcome is termed binocular rivalry. This paper summarizes the conditions triggering rivalry and its accompanying visual consequences, with emphasis on aspects of visual processing that remain functional during monocular suppression.
3-D Stereo Displays: How to Move in the
Third Dimension
T. J. Solz, Jr., K. K. Liggett, J. M.
Reising
Advanced Cockpit Branch, Wright-Patterson AFB, OH
D. C.
Hartsock
Veda, Inc., Dayton, OH full paper
With the possibility of the incorporation of 3-D stereo displays into military cockpits, crew system designers must ensure that the traditional tasks supported by these displays are still possible. This line of research investigated how pilots designated targets on a map display that was portrayed in 3-D stereo. An examination of continuous controllers, aiding algorithms, and discrete controllers is reported.
Depth and Size Judgments within
Uncollimated and Collimated Stereoscopic Displays
R. V.
Parrish
NASA Langely Research Center, Hampton, VA
S. P. Williams
U.S. Army, Langley Research Center, Hampton, VA
D. E. Nold
George
Washington University, Washington, DC full
paper
In collimated stereo displays, the screen is hypothetically placed at infinity. Thus divergent disparity should not place objects any farther away than screen depth (zero disparity). This hypothesis was investigated, as was the influence of different types of display content on depth perception in a collimated stereo system. In a separate uncollimated study, various stereo distance mappings were investigated by asking subjects to adjust the size of a familiar object (a soft drink can) presented at various depths behind the screen until it was the correct size for that distance presented. A reference can was presented at the screen depth.
Time-to-Contact Estimation in a
Driving Simulator
B. Kappé, J. E. Korteling
TNO
Human Factors Research Institute, Soesterberg, The Netherlands
W. A.
Van de Grind
Utrecht University, The Netherlands
full paper
Visual information car drivers use in the perception of the time to contact with an approaching object was studied. Previous research suggested that stationary subjects do not use static size and distance cues when estimating TTC (Schiff & Detwiler, 1979). In addition, an analysis shows that TTC perception involves dynamic information of 'optic flow'. In an experiment in a driving simulator this hypothesis was tested: stationary and moving observers estimated TTC in scenes with a variable amount of cues on TTC. The results show that observers use dynamic information or optic flow when estimation TTC, but they also show that static cues on size and distance are used to reduce a size bias in the estimates.
Visual Requirements for
Virtual-Environment Generation
R. V. Kenyon, T. A.
DeFanti, D. J. Sandin
University of Illinois at Chicago, Chicago, IL full paper
The visual requirements, including field of view, resolution, scene complexity, frame rate, brightness, contrast, distortion, color, focusing, and depth perception, within an immersive synthetic environment (CAVE) that supports full stereoscopic viewing, are addressed
Binocular
Vision in Stereoscopic Virtual-Reality Systems
M.
Mon-Williams, S. Rushton, J. P. Wann
University of Edinburgh,
Scotland full paper
Adverse visual effects may follow use of binocular displays, but can be avoided through careful design and removal of stereoscopic disparity. Display development is discussed in the light of our research into the modification of the reciprocal links between accommodation and vergence when conflict exists between image disparity and focal depth.
Perspective Projection and Its
Tolerances in Partial-Overlap Binocular Head-Mounted Displays
B. H. Tsou
Armstrong Laboratory, Wright-Patterson AFB, OH
S.
S. Grigsby
Logicon Technical Services, Inc., Dayton, OH
R.
Jones, G. E. Pierce
College of Optometry, Ohio State University,
Columbus, OH full paper
Partially overlapping the oculars of binocular head-mounted displays may add perspective distortion to the intrinsic optical distortion. These distortions are both static and dynamic. Static perspective distortion can project false depth cues, while dynamic perspective distortion can introduce false motion cues. Both need to be corrected. Tolerance parameters, as well as a method for correcting the distortions based in part on a survey of the existing clinical literature, are presented
Statistical Study of Eye
Movement When Using a Head-Mounted Display
H. Yoshimatsu
Sony Corp., Tokyo, Japan full paper
Two measures of eye movements were used to quantify visual fatigue caused by prolonged use of a head-mounted display. Data which show that first-order statistics (enthropy and correlation) of microsaccades embedded within conjugate eye movements predict the visual fatigue as experienced by the subject.
Motion Processing of Stereoscopic Depth Information
R.
Patterson, R. Phinney, C. Bowd, M. Donnelly, S. Becker
Washington
State University, Pullman, WA full paper
Research indicates that stereoscopic depth information provides input to motion processing. Velocity discrimination, apparent motion perception, and the induction of adaptation aftereffects are described. Adding stereopsis to visual displays may enhance the perception of object motion, a hypothesis to be tested in future research.
Visual Characteristics of
Cell-Gap Spacers in a PDLC Projector
T. Nakamura, C. Hamamoto, M. Kobayashi, H. Nakamura, N. Harada
Toshiba
Corp., Yokohama, Japan full paper
A polymer-dispersed liquid-crystal (PDLC) display with opaque spacers was developed for high-brightness projectors. The differences in visual characteristics between opaque and conventional transparent spacers was investigated. Opaque spacers were found to have significant advantages over transparent spacers in PDLC projectors.
Color Nameability Surface for Computer-Displayed Colors. I:
Physiological Mechanisms
S. Guest, D. Van Laar
University
of Portsmouth, Portsmouth, U.K. full paper
Simple color-vision theory, colorimetry, and human factors were integrated in order to produce a physiologically plausible model to allow the prediction of colors which are easy or difficult to name. This model is based on the CIE 1976 Uniform Color Space (UCS), which is commonly used in CRT applications.