Memory-Augmented Cellular Automata for Image Analysis.
Abstract
This paper generalizes cellular automata by allowing the memory size associated with each cell to be a function of the input size. In particular, we define a cellular analog to the tape-bounded Turing machine for bounded cellular, pyramid cellular, and parallel/sequential automata. We focus on the case in which each cell has memory size proportional to the logarithm of the input size, showing the increased capabilities of these machines for executing a variety of basic image analysis and recognition tasks. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 01, 1978
- Accession Number
- ADA065328
Entities
People
- Charles R. Dyer
Organizations
- University of Maryland