Number of the records: 1
Scale Sensitivity of Textural Features
- 1.
SYSNO ASEP 0471593 Document Type C - Proceedings Paper (int. conf.) R&D Document Type Conference Paper Title Scale Sensitivity of Textural Features Author(s) Haindl, Michal (UTIA-B) RID, ORCID
Vácha, Pavel (UTIA-B) RIDNumber of authors 2 Source Title Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications: 21st Iberoamerican Congress, CIARP 2016. - Cham : Springer International Publishing, 2017 / Beltran-Castanon C. ; Nystrom I. ; Famili F. - ISBN 978-3-319-52276-0 Pages s. 84-92 Number of pages 8 s. Publication form Print - P Action CIARP 2016 - 21st Iberoamerican Congress 2016 Event date 08.11.2016 - 11.11.2016 VEvent location Lima Country PE - Peru Event type WRD Language eng - English Country DE - Germany Keywords Textural features ; texture scale recognition sensitivity ; surface material recognition ; Markovian illumination invariant features Subject RIV BD - Theory of Information OECD category Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8) R&D Projects GA14-10911S GA ČR - Czech Science Foundation (CSF) Institutional support UTIA-B - RVO:67985556 UT WOS 000418399200011 EID SCOPUS 85013468585 DOI 10.1007/978-3-319-52277-7_11 Annotation Prevailing surface material recognition methods are based on textural features but most of these features are very sensitive to scale variations and the recognition accuracy significantly declines with scale incompatibility between visual material measurements used for learning and unknown materials to be recognized. This effect of mutual incompatibility between training and testing visual material measurements scale on the recognition accuracy is investigated for leading textural features and verified on a wood database, which contains veneers from sixty-six varied European and exotic wood species. The results show that the presented textural features, which are illumination invariants extracted from a generative multispectral Markovian texture representation, outperform the most common alternatives, such as Local Binary Patterns, Gabor features, or histogram-based approaches. Workplace Institute of Information Theory and Automation Contact Markéta Votavová, votavova@utia.cas.cz, Tel.: 266 052 201. Year of Publishing 2018
Number of the records: 1