Texture classification with kernel principal component analysis

Authors
Kim, KI Jung, K Park, SH Kim, HJ
Citation
Ki. Kim et al., Texture classification with kernel principal component analysis, ELECTR LETT, 36(12), 2000, pp. 1021-1022
Citations number
5
Categorie Soggetti
Eletrical & Eletronics Engineeing
Journal title
ELECTRONICS LETTERS
ISSN journal
00135194 → ACNP
Volume
36
Issue
12
Year of publication
2000
Pages
1021 - 1022
Database
ISI
SICI code
0013-5194(20000608)36:12<1021:TCWKPC>2.0.ZU;2-8
Abstract
Kernel principal component analysis (PCA) is presented as a mechanism for e xtracting textural information. Using the polynomial kernel, higher order c orrelations of input pixels can be easily used as features for classificati on. As a result, supervised texture classification can be performed using a neural network.