Pattern recognition based on wavelet moment invariants

Authors
Citation
Xd. Xu et Yh. Zhou, Pattern recognition based on wavelet moment invariants, J INF M W, 19(3), 2000, pp. 215-218
Citations number
9
Categorie Soggetti
Apllied Physucs/Condensed Matter/Materiales Science
Journal title
JOURNAL OF INFRARED AND MILLIMETER WAVES
ISSN journal
10019014 → ACNP
Volume
19
Issue
3
Year of publication
2000
Pages
215 - 218
Database
ISI
SICI code
1001-9014(200006)19:3<215:PRBOWM>2.0.ZU;2-S
Abstract
To get a locally optimum feature set, a feature selection method was presen ted, which combines divergence with Sequential Forward Selection (SFS). Exp erimental results prove that wavelet moment invariants are superior to Zern ike's moment invariants for pattern recognition, especially for classifying seemingly similar objects with subtle difference.