Scientific and Technical Journal


ISSN Print 2221-3937
ISSN Online 2221-3805
In the last decades the amounts of video information increased to such an extent as to make its cataloguing and processing, as well as search and retrieval, as very important and challenging task. Automatic video annotation, i.e. the semantic description of video content, allows solving this task.
In this paper we present an approach for automatic video annotation, based on the ontology, which is created by an expert manually and is used to account for the domain semantics, as well as a Bayesian network, which is built on the base of that ontology and is used to recognize events and situations in video.
We conduct a brief overview and analysis of the existing methods of Bayesian network structure learning and rationalize the necessity to develop a new approach to create the most efficient network structure. The proposed approach uses the branch-and-bound algorithm, which guarantees finding the most optimal network structure at the point of the algorithm’s termination, and structural constraints, that reduce the search space of the possible network structures and reduce the computational costs.
In this paper a network structure is deemed to be most optimal if it has the highest score function value. We use the Bayesian Dirichlet metric as a score function.
The efficiency of the proposed approach is tested in a video annotation information system, using the CAVIAR test video dataset.
1. Zhu F., Shao L., and Lin M., (2013), Multi-view Action Recognition Using Local Similarity Random Forest and Sensor Fusion, Pattern Recognition Letters, Vol. 34, pp. 20 – 24.
2. Kovalenko N.V., and Antoshchuk S.G. Semantiko-Veroyatnostnaya Set’ dlya Opre-deleniya Konteksta Videopotoka v Sistemah Videonabl’udeniya [Semantic-Probabilistic Network for Video Context Recognition in Vid-eo Systems], (2013), Kharkov, Ukraine, East-ern-European Journal of Enterprise Technolo-gies, Vol. 4, No. 4 (64), pp. 15 – 18.
3. Yehezkel R., and Lerner B., (2009), Bayesian Network Structure Learning by Recursive Autonomy Identification, Journal of Machine Learning Research, Vol. 10, pp. 1527 – 1570.
4. Fenz S., Min T.A., and Hudec M., (2009), Ontology-Based Generation of Bayesian Networks, International Conference on Complex, Intelligent and Software Intensive Systems, pp. 712 – 717.
5. Parviainen P., and Koivisto M., (2009), Exact Structure Discovery in Bayesian Networks with Less Space, Proceedings of the 25th Conference on Uncertainty in Artificial Intelligence, pp. 436 – 443.
6. De Campos C.P., Zeng Z., and Ji Q., (2009), Structure Learning of Bayesian Networks Using Constraint, Proceedings of the 26th International Conference on Machine Learning, pp. 113 – 120.
7. CAVIAR. Context aware Vision Using Image-based Active Recognition, (2015), url: http://homepages.
8. Cooper G.F., and Herskovits E.A., (1992), Bayesian Method for the Induction of Probabilistic Networks from Data, Machine Learning, Vol. 9, pp. 309 – 347.
9. Kazi-Tani Y., Lablack A., Ghomari A., and Bilasco M., (2014), Events Detection Using a Video-surveillance Ontology and a Rule-based Approach, ECCV, pp. 299 – 308.
10. Khurana K., and Chandak M.B., (2013), Video Annotation Methodology Based on Ontology for Transportation Domain, International Journal of Advanced Research in Computer Science and Software Engineering, Vol. 3, Iss. 6, pp. 540 – 548.
Last download:
21 Jan 2020

[ © KarelWintersky ] [ All articles ] [ All authors ]
[ © Odessa National Polytechnic University, 2014-2018. Any use of information from the site is possible only under the condition that the source link! ]