Visual element-based Context Detection in Mobile Environment

OData support
Lukovszki Csaba
Department of Telecommunications and Media Informatics

According to some estimation, 686 million smartphone has been sold in the year 1012. That is 38,8% growth compared to 2011. In some developed economies every third person has a smartphone. From the 310 million citizen of the United States 110 million owns a smartphone. Thanks to the fast spread of the new devices, today crowds held mobile phones with appropriate camera and computing power for image processing. These trends enable ordinary users to enjoy the results of the intensive university researches on this filed in the past years.

Smartphones are more efficient devices for some common tasks. Navigation is one of those. Most smartphones have built-in GPS and preinstalled map applications, so many people use them for navigation. There is one huge problem with GPS-based navigation: it isn’t working indoor, when the GPS signals are shielded. Users have the rightful demand for indoor navigation on the same device. Indoor navigation is required in shopping malls, exhibitions, conferences and in large office buildings too. In my thesis, I will examine whether it is possible to build an indoor positioning system based on logo recognition. During my research I tried the mobile image processing technologies and algorithms for logo recognition. I created an Android application and an image collection to test different processes. I sum up my findings in this paper.


Please sign in to download the files of this thesis.