A Novel Approach for Active Event Based Video Summarization Using Foreground Analysis

Rapid growth of no-informative-videos is one of the major concerns of video analytics in recent time. The field like outdoor and indoor surveillance, home, office and shopping mall monitoring produces gigantic volume of no-informative-videos. A novel active event based video summarization is proposed in this research work to make the video analytics more applicable in those fields. Use of adaptive techniques for noise reduction, background modeling, foreground extraction and analysis make the proposed approach more robust towards active event based summarization and indexing. The results on publicly available datasets and a comparative study based on the objectives of the proposed approach with the same of related tesearch works justify the effectiveness of the proposed approach.