Abstract
AN OVERVIEW OF HUMAN OBSERVER CHARACTERISTICS AND THEIR EFFECTON IMAGE TRANSMISSION AND DISPLAYThomas G. Stockham, Jr.Departments of Computer Science and Electrical EngineeringUniversity of UtahSalt Lake City, Utah 84112The science and technology of image transmission and display has evolved primarilyfrom the diciplines of computer science, electrical engineering and physics. Thus, it isonly natural that techniques and attitudes which have developed are characterized by thestyle in which people in these areas approach the subject of images and their trans-mission, display, and processing. For example, one finds many important differencesbetween the methods of television and those of photography. Moreover, an even greatercontrast is found when comparing the methods of electrical engineering, physics, tele-vision and photography with those now understood to be employed by the human eye and thehuman visual system. Specifically, let us contrast the method by which television,photography and the human visual system, represent image information. In television(particularly digital imaging), image values are represented by signals analogous toquantities of light. They are called intensities. In photography, on the other hand,the representing quantities are concentrations of silver or dyes. Consequently, due tothe natural exponentiating laws governing the interaction of light with these media,they are analogous to the logarithm of quantities of light. They are called densities.The human visual system, on the other hand, (while being generally logarithmicallysensitive) moves a large step further away from representation by physical quantities oflight and produces at very early stages in its processing highly modified versions of thepatterns of light or their logarithms. The natural question then arrises, why should thehuman visual system try to do this; and since it does, what consequences are implied interms of the television and photographic presentations normally employed?The answers to these questions rest crucially on the issue of errors. In any trans-mission or display system, errors will be committed. These are unavoidable and are aresult of the physical limitations encountered. In a broad variety of applications, themost important forms of error encountered are those imposed by limited dynamic range andvarious forms of noise. The classic dilemma one faces in the diciplines of image trans-mission and display is that a compromise must be effective between the conflicting con-straints of dynamic range and noise. On the one hand, one wishes to make signals largerso that the noise may be rendered negligible. At the same time, large signals are pre-cluded by the limitation in dynamic range in the form of distortions which effect thesignals when their values are made too variable. More specifically, in any transmissionand /or display design, the natural goal would seem to be the attainment of fidelityreproduction. Unfortunately, the demands of typical imagery upon transmission and dis-play systems is usually so severe that this goal cannot be reasonably met. The dis-tortions due to limited dynamic range and noise, especially the former, are relentlesslyunavoidable and so the designer must content himself with one form of distortion oranother.Fortunately, the human visual system itself produces a large quantity of distortion.This phenomenon may be exploited by trading off undesirable forms for forms which will beencountered naturally anyway. The distortions produced by the human visual system areoften referred to as optical illusions. The simplest of these is the gray scale illusionwhich reveals the logarithmic sensitivity of the human visual system alluded to above(1).This illusion is responsible for the fact that we do not turn the lights on during theday although they add just as much light then as at night. Another ramification is thatgray scales must be arranged in exponential progression to appear arithmetic to theobserver.Other illusions which are spatial in character are much more important and striking,however. The simplest of these is the illusion of simultaneous contrast which permitstwo neutral gray shades to appear so different from one another at the same time, thatone can assign the names near -white to one and near -black to the other. Careful study ofthese illusions has permitted researchers to formulate signal processing models for theearly processing stages of the human visual system. These models (e.g. consider those ofStockham(2), Baudelaire(3), Frei(4), and Baxter(5)) predict the human visual illusions ordistortions well enough to permit their use in defining useful objective measures of howdifferent two images (one original, the other distorted) will look from one another.Applications of these models to problems in image transmission and display havealready yielded significant advantages. For example, when the model is used to evaluatethe distortions produced by a given display instrument, it is often possible to predicthow the physical performance of that display may be significantly relaxed without
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.