Hardware neural network for a visual inspection system

Seungwoo Chun, Yoshihiro Hayakawa, Koji Nakajima

Research output: Contribution to journalArticlepeer-review

Abstract

The visual inspection of deffcts in products is heavily dependent on human experience and instinct. In this situation, it is difficult to reduce the production coast and to shorten the inspection time and hence the total process time. Consequently people involve in this area desire an automatic inspection system In this paper, we propose a hardware neural network which is expected to provide high-speed operation for automatic inspection of products. Since neural network can learn, this is a suitable method for self-adjustmet of criteria for classification. To achievhe high speed operation, we use parallel and pipelining techniques .Furthermore we use a piecewise linear function instead of a conventional activation funtion in order tosave hardware resources. Consequentoly, our proposed hardware neuralnetwork achieve 6GCPS and 2GCUPS, which in our tests sample prove tobe sufficiently fast.

Original languageEnglish
Pages (from-to)935-942
Number of pages8
JournalIEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences
VolumeE91-A
Issue number4
DOIs
Publication statusPublished - 2008 Jan 1

Keywords

  • Back-propagation PCI-BUS
  • FPGA
  • Hardware
  • Visual inspectiosn system

ASJC Scopus subject areas

  • Signal Processing
  • Computer Graphics and Computer-Aided Design
  • Electrical and Electronic Engineering
  • Applied Mathematics

Fingerprint Dive into the research topics of 'Hardware neural network for a visual inspection system'. Together they form a unique fingerprint.

Cite this