Accession Number : ADA331279

Title :   CNN Universal Chip

Descriptive Note : Final rept. 1 Sep 95-31 Aug 97

Corporate Author : CALIFORNIA UNIV BERKELEY ELECTRONICS RESEARCH LAB

Personal Author(s) : Bokor, J. ; Chua, Leon

PDF Url : ADA331279

Report Date : 31 OCT 1997

Pagination or Media Count : 7

Abstract : The goal of designing and building a high performance CNN Universal Machine integrated circuit with demonstrable image processing capabilities was achieved during the contract funding period. The design of the CNN Universal Chip architecture was conducted under the guidance of theoretical results explicating the robust elemental processing capabilities of such cellular analog hardwares, which were developed under this contract. Furthermore, the analogic algorithms needed for implementing the specific image processing techniques required in several target application areas were designed.

Descriptors :   *IMAGE PROCESSING, *CHIPS(ELECTRONICS), *INTEGRATED CIRCUITS, ALGORITHMS, NEURAL NETS, COMPUTER ARCHITECTURE, PARALLEL PROCESSORS, SPATIAL FILTERING, ANALOG SYSTEMS, LOW PASS FILTERS, GRAY SCALE.

Subject Categories : Electrical and Electronic Equipment
      Cybernetics

Distribution Statement : APPROVED FOR PUBLIC RELEASE