Share Email Print
cover

Proceedings Paper

A hardware architecture for fast video object recognition using SVM and Zernike moments
Author(s): Cedric Lemaitre; Johel Miteran; Olivier Aubreton; Romuald Mosqueron
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

An architecture for fast video object recognition is proposed. This architecture is based on an approximation of featureextraction function: Zernike moments and an approximation of a classification framework: Support Vector Machines (SVM). We review the principles of the moment-based method and the principles of the approximation method: dithering. We evaluate the performances of two moment-based methods: Hu invariants and Zernike moments. We evaluate the implementation cost of the best method. We review the principles of classification method and present the combination algorithm which consists in rejecting ambiguities in the learning set using SVM decision, before using the learning step of the hyperrectangles-based method. We present result obtained on a standard database: COIL-100. The results are evaluated regarding hardware cost as well as classification performances.

Paper Details

Date Published: 29 May 2007
PDF: 12 pages
Proc. SPIE 6356, Eighth International Conference on Quality Control by Artificial Vision, 63560X (29 May 2007); doi: 10.1117/12.736745
Show Author Affiliations
Cedric Lemaitre, Lab. Le2i, CNRS, Univ. de Bourgogne (France)
Johel Miteran, Lab. Le2i, CNRS, Univ. de Bourgogne (France)
Olivier Aubreton, Lab. Le2i, CNRS, IUT Le Creusot (France)
Romuald Mosqueron, Lab. Le2i, CNRS, Univ. de Bourgogne (France)


Published in SPIE Proceedings Vol. 6356:
Eighth International Conference on Quality Control by Artificial Vision

© SPIE. Terms of Use
Back to Top