Logo Recognition
|
|
People:
|
|
Abstract
A logo is a graphic entity containing colors, shapes, textures, and perhaps text as well, organized in some spatial layout format. Logo localization and recognition is a subproblem of object detection and recognition and a challenging pattern recognition task. Being of interest for the marketing industry (e.g. to measure the impact of an advertising campaign), trademark registration or vehicle tracking, logo recognition has gained consistent attention in the last few years. Yet, the problem of integrated recognition (i.e. detection/localization + recognition) still remains unresolved.
Papers
- Raluca Boia, Corneliu Florea, "Homographic class template for logo localization and recognition" In IbPRIA 2015, Santiago de Compostela, Spain. [preprint]
- Raluca Boia, Alessandra Bandrabur and Corneliu Florea, "Local Description Using Multi-Scale Complete Rank Transform for Improved Logo Recognition" In Comm 2014, Bucharest, Romania . [preprint] [code]
Databases: FlickrLogos-32 and BelgaLogos
Improved results for integrated recognition (localization and classification):
:
:
