Support Vector Machines With Constraints for Sparsity in the Primal Parameters
A Multi-class Image Classification System Using Salient Features and Support Vector Machines
8
University of Wollongong Research Online Faculty of Informatics - Papers Faculty of Informatics 2007 A multi-class image classification system using salient features and support vector machines Wenbin Shao University of Wollongong, [email protected] Son Lam Phung University of Wollongong, [email protected] G. Naghdy University of Wollongong, [email protected] Research Online is the open access institutional repository for the University of Wollongong. For further information contact Manager Repository Services: [email protected]. Recommended Citation Shao, Wenbin; Phung, Son Lam; and Naghdy, G.: A multi-class image classification system using salient features and support vector machines 2007. http://ro.uow.edu.au/infopapers/561
Transcript of A Multi-class Image Classification System Using Salient Features and Support Vector Machines
University of WollongongResearch Online
Faculty of Informatics - Papers Faculty of Informatics
2007
A multi-class image classification system usingsalient features and support vector machinesWenbin ShaoUniversity of Wollongong, [email protected]
Son Lam PhungUniversity of Wollongong, [email protected]
G. NaghdyUniversity of Wollongong, [email protected]
Research Online is the open access institutional repository for theUniversity of Wollongong. For further information contact ManagerRepository Services: [email protected].
Recommended CitationShao, Wenbin; Phung, Son Lam; and Naghdy, G.: A multi-class image classification system using salient features and support vectormachines 2007.http://ro.uow.edu.au/infopapers/561
A multi-class image classification system using salient features and supportvector machines
AbstractThis paper addresses the problem of automatic image annotation for semantic retrieval of images. Wepropose an image classification system that is capable of recognizing several image categories. The system isbased on the support vector machine and a set of image features that includes MPEG-7 visual descriptors anda custom feature. The system is evaluated on a large dataset consisting of 14400 images in four categories -landscape, cityscape, vehicle and portrait. We find that the proposed edge direction histogram and theMPEG-7 edge histogram perform better than other features in this application. Experiment results indicatethat the pair- wise SVM approach performs better than the one-versus-all SVM approach. The pair-wisemethod with confidence score voting has better classification rates compared to the pair-wise method withmajority voting.
Keywordsimage classification, MPEG-7, SVM, image annotation, image retrieval
Publication DetailsThis paper was originally published as Shao, W, Phung, SL and Naghdy, G, A multi-class image classificationsystem using salient features and support vector machines, Proceedings of International Conference onSensors, Sensor Networks and Information Processing (ISSNIP 2007), Melbourne, Australia, 3-6 December2007, 431-436. Copyright 2007 IEEE.
This conference paper is available at Research Online: http://ro.uow.edu.au/infopapers/561
