Logotype Detection in Streaming Multimedia Using Apache Storm

José Herrera, Germán Moltó, and Atsuhiro Takasu. Logotype Detection in Streaming Multimedia Using Apache Storm. In Proceedings of the 2017 International Conference on Parallel and Distributed Processing Techniques and Applications (PDPTA), pp. 81–87, 2017.

Download

[HTML] 

Abstract

The massive amount of multimedia information currently available through the Internet requires efficient techniques to extract knowledge from data. Indeed, processing video information using real-time Big Data techniques is currently a challenge for the existing data processing frameworks. This paper proposes and evaluates different topologies for logotype detection in streaming multimedia implemented by means of open-source libraries and supported by the Apache Storm distributed real-time computation system. Different data movement strategies across the topologies are defined and evaluated. An experimental analysis that involves logotype detection in an excerpt of the Formula One Suzuka Japan Grand Prix 2015 is performed. The results indicate that logotype detection in videos can greatly benefit from the distributed computing capabilities offered by Apache Storm.

BibTeX Entry

@inproceedings{Herrera2017lds,
   abstract = {The massive amount of multimedia information currently available through the Internet requires efficient techniques to extract knowledge from data. Indeed, processing video information using real-time Big Data techniques is currently a challenge for the existing data processing frameworks. This paper proposes and evaluates different topologies for logotype detection in streaming multimedia implemented by means of open-source libraries and supported by the Apache Storm distributed real-time computation system. Different data movement strategies across the topologies are defined and evaluated. An experimental analysis that involves logotype detection in an excerpt of the Formula One Suzuka Japan Grand Prix 2015 is performed. The results indicate that logotype detection in videos can greatly benefit from the distributed computing capabilities offered by Apache Storm.},
   author = {José Herrera and Germán Moltó and Atsuhiro Takasu},
   booktitle = {Proceedings of the 2017 International Conference on Parallel and Distributed Processing Techniques and Applications (PDPTA)},
   keywords = {Apache Storm,Big Data,Bloom Filter,logotypes,multimedia},
   pages = {81-87},
   title = {Logotype Detection in Streaming Multimedia Using Apache Storm},
   url = {http://csce.ucmss.com/books/LFS/CSREA2017/PDP3398.pdf},
   year = {2017}
}