Marr's main work was on three topics: first, the method by which the visual system recovers lines and edges, which, although they look so clean to the viewer, are extremely hard to reconstruct from the messy retinal image; second, stereopsis, particularly the correspondence problem, that is the method by which the brain matches a point in the image on one eye with the equivalent point in the other eye's image; and, third, the problem of how objects are represented in the brain in such a way as to facilitate recognition. Although he made highly original and important contributions to all three problems, he will probably be remembered less for the details of his research than for the style of his approach. Before Marr, most computer programs in vision dealt only with very restricted 'toy' worlds, for example plane-sided blocks. Marr set out to model vision as it operates in the real world in all its complexity. Moreover, he was the first person to construct computer models that took account of neurophysiological and psychological findings. He did not want to produce just any model — he wanted to uncover the way the human visual system actually works. Finally, he saw more clearly than almost any of his contemporaries that, in order to build an adequate model of how the brain executes a certain task, it is first necessary to obtain a clear understanding of that task: his mathematical ability helped him to achieve this kind of understanding. In addition to carrying out his own research, Marr through his intelligence, charm, and liveliness attracted around him a group of outstanding research workers drawn from all over the world. Inspired by him, they made important contributions of their own, many of which are recorded in his book.
David Marr's move to MIT seemed to lead to a remarkable flowering of his imagination and personality. Part of his vitality over the last few years of his life may have come from the fact that he knew he had not long to live. Three years before he died he contracted leukaemia, a disease that necessitated repeated and painful stays in hospital. He faced his illness with quiet courage and, no matter how weak he felt, he went on working: his fascination with his work was so great that however ill he was he never had to force himself to do it. His attempt to grapple with the deepest problems of vision was for him a source of spontaneous enjoyment. As he writes at the beginning of his book, 'This book is meant to be enjoyed'. One can be sure he enjoyed writing it.
(Published 1987)
See also perception.
— Stuart Sutherland
