STATISTICALLY CHARACTERIZED FEATURES FOR DIRECTIONALITY QUANTITATION IN PATTERNS AND TEXTURES

Authors
DENSLOW S ZHANG Z THOMPSON RP LAM CF
Citation
S. Denslow et al., STATISTICALLY CHARACTERIZED FEATURES FOR DIRECTIONALITY QUANTITATION IN PATTERNS AND TEXTURES, Pattern recognition, 26(8), 1993, pp. 1193-1205
Citations number
26
Categorie Soggetti
Computer Sciences, Special Topics","Engineering, Eletrical & Electronic","Computer Applications & Cybernetics
Journal title
Pattern recognition → ACNP
ISSN journal
00313203
Volume
26
Issue
8
Year of publication
1993
Pages
1193 - 1205
Database
ISI
SICI code
0031-3203(1993)26:8<1193:SCFFDQ>2.0.ZU;2-X
Abstract
A method is presented for numerical characterization of directionality in biological images by extraction of angle and strength features fro m image areas. The usefulness of these features is expanded by (1) the documentation of the functional behavior of the heuristically defined strength features and (2) the derivation and testing of the statistic al properties of angle and strength of directionality.