Automatic TV logo identification

Author

Özay, Nedret ; Sankur, Bülent

fYear

2008

fDate

20-22 April 2008

Firstpage

1

Lastpage

4

Abstract

In this work, a TV logo identification system has been developed. In preprocessing step all logos are scaled from pixel-based representation to a fixed size of macro-pixel representation. Commonly used techniques in pattern recognition, PCA (principal components analysis), NMF (non-negative matrix factorization), DCT (discrete cosine transform), and ICA (independent component analysis) are used to extract features in logos. Those feature vectors then used in SVM (support vector machines) for classification. Both colour and gray-scale images were used and importance of colour in logo identification has been observed. Accuracy rate achieved on a test set of 760 logos is 99.21% for colour images.

Keywords

discrete cosine transforms; feature extraction; image classification; image colour analysis; image recognition; image representation; image resolution; independent component analysis; matrix decomposition; principal component analysis; support vector machines; DCT; ICA; PCA; SVM; automatic TV logo identification; classification; colour images; discrete cosine transform; feature vectors; independent component analysis; macropixel representation; nonnegative matrix factorization; pattern recognition; pixel-based representation; principal components analysis; support vector machines; Discrete cosine transforms; Feature extraction; Gray-scale; Independent component analysis; Pattern recognition; Principal component analysis; Support vector machine classification; Support vector machines; TV; Testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Signal Processing, Communication and Applications Conference, 2008. SIU 2008. IEEE 16th

Conference_Location

Aydin

Print_ISBN

978-1-4244-1998-2

Electronic_ISBN

978-1-4244-1999-9

Type

conf

DOI

10.1109/SIU.2008.4632632

Filename

4632632