Combined statistical and model based texture features for improved image classification

Al-Kadi, Omar S (2008) Combined statistical and model based texture features for improved image classification. 4th International Conference on Advances in Medical, Signal & Information Processing, Santa Margherita Ligure, Italy, 14-16 July 2008. Published in: Proceedings of 4th IET International Conference on Advances in Medical, Signal and Information Processing; Italy; 14-16 July 2008. Institute of Electrical and Electronics Engineers ISSN 0537-9989 ISBN 9780863419348

[img] PDF (2008 IEEE) - Accepted Version
Download (638kB)

Abstract

This paper aims to improve the accuracy of texture classification based on extracting texture features using five different texture measures and classifying the patterns using a naive Bayesian classifier. Three statistical-based and two model-based methods are used to extract texture features from eight different texture images, then their accuracy is ranked after using each method individually and in pairs. The accuracy improved up to 97.01% when model based - Gaussian Markov random field (GMRF) and fractional Brownian motion (fBm) - were used together for classification as compared to the highest achieved using each of the five different methods alone; and proved to be better in classifying as compared to statistical methods. Also, using GMRF with statistical based methods, such as grey level co-occurrence (GLCM) and run-length (RLM) matrices, improved the overall accuracy to 96.94% and 96.55%; respectively.

Item Type: Conference Proceedings
Additional Information: 2008 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collect.
Depositing User: Omar Al-Kadi
Date Deposited: 21 Feb 2012 10:01
Last Modified: 03 May 2018 13:15
URI: http://sro.sussex.ac.uk/id/eprint/37622

View download statistics for this item

📧 Request an update