Computing motion using resistive networksRecent developments in the theory of early vision are described which lead from the formulation of the motion problem as an ill-posed one to its solution by minimizing certain 'cost' functions. These cost or energy functions can be mapped onto simple analog and digital resistive networks. It is shown how the optical flow can be computed by injecting currents into resistive networks and recording the resulting stationary voltage distribution at each node. These networks can be implemented in cMOS VLSI circuits and represent plausible candidates for biological vision systems.
Document ID
19890027111
Acquisition Source
Legacy CDMS
Document Type
Conference Paper
Authors
Koch, Christof (California Inst. of Tech. Pasadena, CA, United States)
Luo, Jin (California Inst. of Tech. Pasadena, CA, United States)
Mead, Carver (California Institute of Technology Pasadena, United States)
Hutchinson, James (California Institute of Technology Jet Propulsion Laboratory, Pasadena, United States)
Date Acquired
August 14, 2013
Publication Date
January 1, 1988
Subject Category
Cybernetics
Meeting Information
Meeting: Neural Network Models for Optical Computing