Local description using multi-scale complete rank transform for improved logo recognition

Author

Boia, Raluca ; Bandrabur, Alessandra ; Florea, Catalin

Author_Institution

Image Process. & Anal. Lab. (LAPI), Univ. “Politeh.”, Bucharest, Romania

fYear

2014

fDate

29-31 May 2014

Firstpage

1

Lastpage

4

Abstract

In this paper we address the challenging problem of logo recognition. As logos appear in various scales, colors and with illuminations coming from source lights of various intensities, positions or colors, their recognition is a tedious task. The typical approach stands within Bag-of-Words (BoW) framework with SIFT (Scale Invariant Feature Transform) based features. To increase the accuracy, we rely on the recently introduced feature of the complete rank transform, which we extend for a multi-scale description. The system´s performance is verified on the FlickrLogos-32 database.

Keywords

image recognition; transforms; visual databases; BoW; FlickrLogos-32 database; SIFT; bag-of-words framework; local image descriptors; logo recognition; multiscale complete rank transform; scale invariant feature transform; Accuracy; Databases; Feature extraction; Image color analysis; Transforms; Visualization; Vocabulary; local image descriptors; logo recognition; object recognition;

fLanguage

English

Publisher

ieee

Conference_Titel

Communications (COMM), 2014 10th International Conference on

Conference_Location

Bucharest

Type

conf

DOI

10.1109/ICComm.2014.6866723

Filename

6866723