David Marr, 1982


Part I. Introduction and Philosophical Preliminaries

General Introduction
What does it mean to see? The plain man's answer (and Aristotle's, too) would be, to know what is where by looking. In other words, vision is the process of discovering from images what is present in the world, and where it is.

Vision is therefore, first and foremost, an information-processing task, but we cannot think of it as just a process. For if we are capable of knowing what is where in the world, our brains must somehow be capable of representing this information -- in all its profusion of color and form, beauty, motion, and detail. The study of vision must therefore include not only the study of how to extract from images the various aspects of the world that are useful to us, but also an inquiry into the nature of the internal representations by which we capture this information and thus make it available as a basis for decisions about our thoughts and actions.

Once people began to dream of information-processing tasks and then to build such machines, it soon became clear that many aspects of the world around us could benefit from an information-processing point of view. Most of the phenomena that are central to us as human beings -- the mysteries of life and evolution, of perception and feeling and thought -- are primarily phenomena of information processing, and if we are ever to understand them fully, our thinking about them must include this perspective.

To say that a job is "only" an information-processing task or that an organism is "only" an information-processing machine is not a limiting or a perjorative description. Quite the contrary, in fact. One of the fascinating features of information-processing machines is that in order to understand them completely, one has to be satisfied with one's explanations at many different levels.

Think, for example, of the international network of airline reservation computers, which performs the task of assigning flights for millions of passengers all over the world. To understand this system, it is not enough to know how a modern computer works. One also has to understand a little about what aircraft are and what they do; about geography, time zones, fares, exchange rates and connections; and something about politics, diets and the various other aspects of human nature that happen to be relevant to this particular task.

To understand fully a particular machine carrying our a particular information-processing task, one has to study the computer and the information-processing task. Neither alone will suffice.

Chapter 1. The Philosophy and the Approach


Understanding Complex Information-Processing Systems

A Representational Framework for Vision

Part II. Vision

Chapter 2. Representing the Image

Physical Background of Early Vision

Zero-Crossings and the Raw Primal Sketch

Spatial Arrangement of an Image

Light Sources and Transparency

Grouping Processes and the Full Primal Sketch

Chapter 3. From Images to Surfaces

Modular Organization of the Human Visual Processor

Processes, Constraints, and the Available Representations of an Image


Directional Selectivity

Apparent Motion

Shape Contours

Surface Texture

Shading and Photometric Stereo

Brightness, Lightness, and Color


Chapter 4. The Immediate Representation of Visible Surfaces


Image Segmentation

Reformulating the Problem

The Information to be Represented

General Form of the 2 1/2 D Sketch

Possible Forms for the Representation

Possible Coordinate Systems

Interpolation, Continuation, and Discontinuities

Computational Aspects of the Interpolation Problem

Other Internal Computations

Chapter 5. Representing Shapes for Recognition


Issues Raised by the Represenation of Shape

The 3-D Model Representation

Natural Extensions

Deriving and Using The 3-D Model Representation

Psychological Considerations

Chapter 6. Synopsis


Part III. Epilogue

Chapter 7. A Conversation


A Way of Thinking

Top of Page | V Opinion | Sort by Topic | Sort by Title | Sort by Author