Bayesian System and Copula for Event Detection and Summarization of Soccer Videos

Title: Bayesian System and Copula for Event Detection and Summarization of Soccer Videos
Publisher: Guru Nanak Publications
ISSN: 2278-0947
Series: Volume 4 Issue 2
Authors: Dhanuja S. Patil and Prof. Sanjay B. Waykar


Abstract

Event detection is a standout amongst the most key parts for distinctive sorts of area applications of video data framework. Recently, it has picked up an extensive interest of experts and in scholastics from different zones. Recently, while detecting video event has been the subject of broad study efforts, impressively less existing methodology has considered multi-model data and issues related efficiency. Start of soccer matches different doubtful circumstances grow that can't be effectively judged by the referee committee. A framework that verifies objectively image arrangements would prevent not right interpretations because of some errors, or high velocity of the events. Bayesian networks give a structure for dealing with this vulnerability using an essential graphical structure likewise the probability analytics. We propose an efficient structure for analyzing and summarization of soccer videos utilizing object-based features. The proposed work utilizes the t-cherry junction tree, an exceptionally recent advancement in probabilistic graphical models, to create a compact representation and great approximation intractable model for client’s relationships in an interpersonal organization. Examination results demonstrates the effectiveness, adequacy, and the strength of the proposed work which is shown over a far reaching information set, comprising more soccer feature, caught at better places.

Keywords

Summarization, detection, Bayesian network, t-cherry tree.

Download Full Text

(For complimentary copy, please contact Chief_editor@innovationjournals.com)