variations to which the image is prey (such as lighting effects), and thus will be a more reliable basis for practical decisions such as manipulation, navigation, or recognition. Chapters 1.2 and 1.3 are mainly concerned with intrinsic images calculation, using clues of shading, multiple cameras, and motion.
System Architecture . System architectures for vision are now being designed as part of several major projects in research centers around the country. Often the system is built around a "blackboard" that serves as communications medium, central representation, and sometimes as a form of autonomous inference or geometrical computing engine. In this volume, chapter 1 gives an explicit description of the detailed makeup of such a vision system. The neural architecture that could underlie vision and thinking in general is of course an important topic in the cognitive sciences. Chapter 2.3 outlines a hierarchy of abstractions implemented in a neural-like net of small processing units.
Hardware Architecture . There is much activity today in designing and building special-purpose hardware architectures. Low-level vision (like graphics and image processing) has long been a candidate for such work because of the simplicity and physical locality of many of the operations. Recently high-level vision, or cognitive processes in general, has influenced the development of computational models (chapter 2.3; Feldman, 1985) and actual computers ( Hillis, 1986). Advances in the power and packaging of microcomputers have led to new general-purpose architectures involving many computers working in parallel. These and more special-purpose hardware architectures are very promising tools for advancing all levels of the computer vision problem. In this volume the hardware architecture issue is addressed in chapter 1.
Ballard, D., & Brown, C. ( 1982). Computer vision. Englewood Cliffs, NJ: Prentice-Hall.
Feldman, J. A. Editor. (January-March 1985). Cognitive Science 9, 1, Special Issue on Connectionist Models.
Fodor, J. ( 1983). The modularity of mind. Cambridge, MA: MIT Press.
Hillis, D. ( 1985). The connection machine. Cambridge, MA: MIT Press.
Horn, B. K. P. ( 1986). Robot vision. Cambridge, MA: MIT Press.
Levine, M. ( 1985). Vision in man and machine. New York: McGraw-Hill.
Marr, D. ( 1982). Vision. San Francisco, CA: Freeman.
Pentland, A. P. ( 1986). From pixels to predicates. Norwood, NJ: Ablex.
Roberts, L. G. ( 1965). "Machine perception of three-dimensional solids". In J. P. Tippet et al. (Eds.), Optical and electro-optical information processing (pp. 159-197). Cambridge, MA: MIT Press.