Using distributed neural networks in automated optical inspection

Author

Milea, Constantin ; Svasta, Paul

Author_Institution

Bucharest Politehnica Univ., Romania

fYear

2001

fDate

2001

Firstpage

286

Lastpage

288

Abstract

This paper introduces the concept of distributed neural networks over a computer network. Each layer in the neural network can be run as a separate application on different computers or on the same computer. The paper emphasizes the advantages of such an architecture over classical ones in terms of issues such as large neural networks and large input data. To test this new architecture, we use images from some automated optical inspection machinery, hoping to improve the results

Keywords

automatic optical inspection; computer networks; image processing; neural net architecture; automated optical inspection; automated optical inspection machine images; computer network; distributed neural networks; input data; large neural networks; network architecture; neural network layers; separate application layers; Application software; Automatic optical inspection; Automatic testing; Computer applications; Computer architecture; Computer networks; Machinery; Neural networks; Optical computing; Optical fiber networks;

fLanguage

English

Publisher

ieee

Conference_Titel

Electronics Technology: Concurrent Engineering in Electronic Packaging, 2001. 24th International Spring Seminar on

Conference_Location

Calimanesti-Caciulata

Print_ISBN

0-7803-7111-9

Type

conf

DOI

10.1109/ISSE.2001.931084

Filename

931084