DocumentCode :
594708
Title :
Impact of a codebook filtering step on a galois lattice structure for graphics recognition
Author :
Boumaiza, A. ; Tabbone, Salvatore
Author_Institution :
LORIA, Univ. de Lorraine, Vandœuvre-lès-Nancy, France
fYear :
2012
fDate :
11-15 Nov. 2012
Firstpage :
278
Lastpage :
281
Abstract :
In this paper, we propose an evaluation of the impact of a codebook filtering step on the recognition rate of a Galois lattice classifier. Unlike the usual approach which only considers a whole visual dictionary and is likely to over-fitting, we boost the Galois lattice using a filtered dictionary by assigning a probability of appearance to each visual word in a symbol model. The retrieval performance and behavior of the method have been compared to state-of-the art and proved that is suitable to the recognition process. Experimental results show that the Galois Lattice classifier combined with a filtered codebook outperforms classic classifiers. Interestingly, due to the high selection of features from the dictionary, the accuracy improvement is obtained with a considerable computational cost reduction.
Keywords :
Galois fields; computer graphics; feature extraction; image classification; image retrieval; information filtering; probability; Galois lattice classifier; Galois lattice structure; codebook filtering; computational cost reduction; feature selection; filtered dictionary; graphics recognition; over-fitting; retrieval performance; symbol model; visual dictionary; visual word appearance probability; Buildings; Dictionaries; Filtering; Lattices; Visualization; Vocabulary;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Pattern Recognition (ICPR), 2012 21st International Conference on
Conference_Location :
Tsukuba
ISSN :
1051-4651
Print_ISBN :
978-1-4673-2216-4
Type :
conf
Filename :
6460126
Link To Document :
بازگشت