Ординатура / Офтальмология / Английские материалы / Seeing_De Valois_2000
.pdfx Contents |
|
D. Preattentive Features |
344 |
E. The Preattentive Processing of Objects |
354 |
F. Preattentive Summary |
358 |
III. Vision with Attention |
358 |
A. Attention Enables Other Visual Processes |
358 |
B. How and What Does Attention Enable? |
358 |
IV. Vision after Attention |
364 |
A. Repeated Search |
365 |
B. Change Blindness |
366 |
V. Vision without Attention |
367 |
A. The Problem |
367 |
B. How Unattended Can You Get? |
368 |
C. All Inattention Is Not Created Equal |
369 |
VI. Conclusion |
369 |
References |
370 |
Index |
387 |
Contributors
Numbers in parentheses indicate the pages on which the authors’ contributions begin.
Duane G. Albrecht (79)
Psychology Department
University of Texas
Austin, Texas 78712
Andrew Derrington (259)
School of Psychology
University of Nottingham
Nottingham NE7 2RD, England
Karen K. De Valois (129)
Departments of Psychology and Vision
Science
University of California at Berkeley
Berkeley, California 94720
Russell L. De Valois (129)
Departments of Psychology and Vision
Science
University of California at Berkeley
Berkeley, California 94720
Jack L. Gallant (311)
Department of Psychology
University of California at Berkeley
Berkeley, California 94720
Wilson S. Geisler (79)
Department of Psychology
University of Texas
Austin, Texas 78712
Clifton M. Schor (177)
School of Optometry
University of California at Berkeley
Berkeley, California 94720
Robert Shapley (55)
Center for Neural Science
New York University
New York, New York 10003
Larry N. Thibos (1)
School of Optometry
Indiana University
Bloomington, Indiana 47405
Jeremy M. Wolfe (335) Harvard Medical School and Center for Ophthalmic Research Brigham and Women’s Hospital Boston, Massachusetts 02115
xi
This Page Intentionally Left Blank
Foreword
The problem of perception and cognition is in understanding how the organism transforms, organizes, stores, and uses information arising from the world in sense data or memory. With this definition of perception and cognition in mind, this handbook is designed to bring together the essential aspects of this very large, diverse, and scattered literature and to give a précis of the state of knowledge in every area of perception and cognition. The work is aimed at the psychologist and the cognitive scientist in particular, and at the natural scientist in general.Topics are covered in comprehensive surveys in which fundamental facts and concepts are presented, and important leads to journals and monographs of the specialized literature are provided. Perception and cognition are considered in the widest sense. Therefore, the work treats a wide range of experimental and theoretical work.
The Handbook of Perception and Cognition should serve as a basic source and reference work for those in the arts or sciences, indeed for all who are interested in human perception, action, and cognition.
Edward C. Carterette and Morton P. Friedman
xiii
This Page Intentionally Left Blank
Preface
Of all the things that humans can do, seeing is arguably what we do best. The eyes encode and transmit enough information to allow the brain to construct a vivid representation of the three-dimensional structure of the external world, including estimates of the relative intensity and spectral composition of the light at every visible point, all within a fraction of a second. We detect and identify objects, and determine where they are and the direction and speed with which they are moving, with extraordinary sensitivity and precision. The most remarkable thing about seeing, however, is that it appears to be effortless. We are rarely aware of the complex computational tasks our visual systems are accomplishing, and we do not experience seeing as difficult. When we open our eyes, the world simply appears. The apparent ease with which we accomplish vision makes it difficult to recognize just how complicated the visual system is.
In this volume, several specific aspects of seeing are considered. Their selection was based in part on the state of our current knowledge. Some are classical topics, for example, color vision, spatial vision, binocular vision, and visual receptive fields for which there is a broad and deep pool of knowledge. Others (motion vision, image formation, and sampling) are traditional problems about which there has been a relatively recent explosion of new information and understanding. Still others (neural representation of shape, visual attention) have more recently become prominent, in part because only now do we have the tools necessary to begin to understand the mechanisms responsible. Each of these topics could be—indeed, has been—the subject of dedicated volumes. Here we attempt to describe some of the
xv
xvi Preface
problems attendant upon each topic, to give a useful overview of the current state of our knowledge, and to introduce the interested reader to the modern literature.
I am indebted to my colleagues who wrote the various chapters of this book and to the publisher and series editors who decided to bring out a new, expanded version of the Handbook of Perception, now the Handbook of Perception and Cognition. I am grateful to the various authors and publishers who have given permission for the reprinting of figures from earlier publications. They are individually acknowledged and credited where the figures appear. My friends—colleagues, staff, and stu- dents—at Berkeley have been immensely helpful and supportive, and I am grateful. Finally, Russell De Valois has been deeply involved at every stage—writing, reading, editing. Without his help and support, this undertaking would never have been completed.
Karen K. De Valois
C H A P T E R 1
Formation and Sampling
of the Retinal Image
Larry N. Thibos
I. INTRODUCTION
Vision begins with the formation of an optical image of the external world upon the retinal mosaic of light-sensitive photoreceptors. Because image formation is the very first step in the visual process, imperfections in the eye’s optical apparatus have the potential for a ecting every aspect of visual perception, from color to motion, space, and form. Even if the focusing components of the eye’s optical system were perfect, the retinal image would still be degraded by the di raction of light as it passes through the pupil. Another potential limiting factor for vision is the scattering of light as it traverses the ocular media and retina before being absorbed by the visual pigment molecules inside photoreceptors. Thus the study of visual perception must logically begin with the imaging properties of the eye to uncover those optical factors which constrain the quality of vision.
The second step in the visual process is the sampling of the optical image by the retina, a thin layer of neural tissue at the back of the eye. The light-sensitive rod and cone photoreceptors are the transducers of the retina that convert the optical image into a discrete, neural representation of the retinal image. This discrete “neural image” is then resampled by an array of interneurons (called bipolar cells), and this intermediate neural image is then resampled yet again by the output cells of the retina (called ganglion cells) for transmission through the optic nerve to the brain. These multiple sampling operations are of fundamental importance to vision
Seeing
Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.
1
2 Larry N. Thibos
because they limit the fidelity of the neural representation of the external world. Just as a coarse layer of light-sensitive crystals on photographic film produces a coarse picture, so too will a coarse array of light-detecting neural elements represent a visual image coarsely. Consequently, it is physically impossible for a sparse neural array, such as exists in the peripheral parts of the retina, to faithfully represent fine visual patterns. This is not to say that fine patterns fail to produce a useful neural image. To the contrary, it is possible for fine patterns to generate visual signals which yield viable percepts, but these percepts must necessarily misrepresent the stimulus as a coarse pattern.This phenomenon, called aliasing, represents an irretrievable loss of fidelity in the neural image that has the potential for a ecting many aspects of visual perception.
The aim of this chapter is to succinctly review our current understanding of the formation and sampling of the retinal image in the human eye and the visual limitations imposed by these initial stages of the visual process. For the benefit of readers unfamiliar with the physical and mathematical description of image formation, a brief review is provided of the main optical concepts needed to understand how the retinal image is formed and how the quality of that image is characterized. These concepts are then embodied in an optical model of the eye which is useful for qualitative as well as quantitative thinking about retinal images. Similarly, the mathematical concepts needed to formulate the intuitive notion of neural sampling limits on vision are summarized before presenting a model of the sampling process. When taken together, this neuro-optical model of the eye provides a conceptual framework for interpreting empirical studies of the optical and sampling limits to visual perception and the impact these two factors have in daily life.
II. FORMATION OF THE RETINAL IMAGE
A. Optical System of the Eye
The optical components of the adult human eye are shown in anatomical crosssection in Figure 1 (Walls, 1942). In its resting state, two-thirds of the optical power of the eye is provided by refraction of light by the cornea and the remaining third is provided by the internal lens of the eye. However, the internal lens of the eye is capable of changing shape to provide additional focusing power when looking at close objects. An eye is said to be emmetropic if it produces a clearly focused retinal image of distant targets when the lens has minimum power. At birth the human eye is small, but it grows rapidly over the first 3 years of life, and by 3 years of age it is nearly adult size (Bennett & Rabbetts, 1989). During this period a coordinated growth process called emmetropization aims to keep the eye clearly focused for distant targets when the eye is in its resting state (Gwiazda, Thorn, Bauer, & Held, 1993; Wallman, Adams, & Trachtman, 1981). Although the mechanism of emmetropization is not fully understood, it appears to function independently of
1 Retinal Image Formation and Sampling |
3 |
FIGURE 1 An anatomically correct drawing of a human eye in cross-section. (Redrawn from Walls, 1942.)
the brain, since cutting of the optic nerve in experimental animals does not alter the pattern of eye growth (Raviola & Wiesel, 1985; Troilo, Gottlieb, & Wallman, 1987).
During childhood and early adulthood the internal lens of the eye is flexible so its anterior surface can bulge forward to make the lens more powerful optically, thus focusing the eye on closer targets. These changes in shape occur as tension on the supporting zonule fibers is released in response to forward movement of the ciliary processes caused by contraction of the ciliary muscle (Glasser & Campbell, 1998; Helmholtz, 1909/1924). This ability of the human eye to increase its focal power, called accommodation, provides an operating range for focusing that extends from infinity to less than 10 cm from the eye in a young person. The physical unit of focusing power is the diopter (D), which is computed as the inverse of the distance (in meters) from the eye to a target which is clearly focused on the retina. The