MVS: A multi-view video synopsis framework
详细信息 查看全文
- 作者：Ansuman Mahapatraa ; helloansuman@gmail.com" class="auth_mail" title="E-mail the corresponding author ; Pankaj K. Saa ; pankajksa@nitrkl.ac.in" class="auth_mail" title="E-mail the corresponding author ; Banshidhar Majhia ; bmajhi@nitrkl.ac.in" class="auth_mail" title="E-mail the corresponding author ; Sudarshan Padhyb ; spadhy07@gmail.com" class="auth_mail" title="E-mail the corresponding author
- 关键词：Video synopsis ; Multi-view video ; Multi-camera network ; Video summarization
- 刊名：Signal Processing: Image Communication
- 出版年：2016
- 出版时间：March 2016
- 年：2016
- 卷：42
- 期：Complete
- 页码：31-44
- 全文大小：5556 K
文摘
- •
A framework to generate one synopsis of multi-view videos is proposed.
- •
The synopsis is visualized by mapping each camera view to a common ground plane.
- •
The generated synopsis also incorporates the advantages of video summarization.
- •
The action recognition module helps limiting the synopsis length significantly.
- •
Visibility scores of each object track further reduce the synopsis length.