| Authors: | Žunić, Joviša Rosin, Paul Kopanja, Lazar |
Title: | Shape orientability | Journal: | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Volume: | 3852 LNCS | First page: | 11 | Last page: | 20 | Conference: | 7th Asian Conference on Computer Vision, ACCV 2006; Hyderabad; India; 13 January 2006 through 16 January 2006 | Issue Date: | 1-Jan-2006 | Rank: | M23 | ISBN: | 978-3-540-32432-4 | ISSN: | 0302-9743 | DOI: | 10.1007/11612704_2 | Abstract: | In this paper we consider some questions related to the orientation of shapes. We introduce as a new shape feature shape orientability, i.e. the degree to which a shape has distinct (but not necessarily unique) orientation. A new method is described for measuring shape orientability, and has several desirable properties. In particular, unlike the standard moment based measure of elongation, it is able to differentiate between the varying levels of orientability of n-fold rotationally symmetric shapes. |
Publisher: | Springer Link |
Show full item record
SCOPUSTM
Citations
10
checked on Dec 26, 2024
Page view(s)
19
checked on Dec 26, 2024
Google ScholarTM
Check
Altmetric
Altmetric
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.