Abstract
The primary goal of this thesis was to emulate the function of the biological eye in silicon. In both neural and silicon technologies, the active devices occupy approximately 2 percent of the space, wire fills the entire remaining space. The silicon retina was modeled on the distal portion of the vertebrate retina. This chip generates, in real time, outputs that correspond directly to signals observed in the corresponding levels of the biological retinas. The design uses the principles of signal aggregation. It demonstrates a tolerance for device imperfection that is characteristic of a collective system. The digital computer is extremely effective at producing precise answers to well-defined questions. The nervous system accepts fuzzy, poorly conditioned input, performs a computation that is ill-defined, and produces approximate output.
Library of Congress Subject Headings
Neural networks (Computer science); Computer vision; Image processing; Retina--Computer simulation
Publication Date
2-1-1994
Document Type
Thesis
Department, Program, or Center
Computer Engineering (KGCOE)
Advisor
Brown, George
Recommended Citation
Golwalla, Arif K., "A Hardware approach to neural networks silicon retina" (1994). Thesis. Rochester Institute of Technology. Accessed from
https://repository.rit.edu/theses/4617
Campus
RIT – Main Campus
Comments
Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: QA76.87.G64 1994