Possibilities for blazon and logo detection

OData support
Supervisor:
Lukovszki Csaba
Department of Telecommunications and Media Informatics

With the spread of mobile devices and the increasing performance of smartphones the processing of visual information is becoming more and more available on mobile platforms. This provides several possibilities of usage for example painting, face, building, landscape, road sign, poster, location, money, and many other object recognition.

The aim of my thesis is to show a way of logo recognition on Android platform. I present some algorithms of object recognition. Then I verify its applicability on logos. To achieve this an average logo is transformed in a way that can occur in reality (e.g. rotation and scale) and I analyze the success of recognition with tests. In the next step I test the recognition on real-world photos. Then I present solutions to recognize logos modified in different ways. In order to find out its usability on mobile platform I show the main steps of the development of the chosen algorithm on Android. Finally, I present the results I got running the application.

Downloads

Please sign in to download the files of this thesis.