Edge-centric Video Surveillance System Based on Event-driven Rate Adaptation for 24-hour Monitoring

Airi Sakaushi, Kenji Kanai, Jiro Katto, Toshitaka Tsuda

研究成果: Conference contribution

3 被引用数 (Scopus)

抄録

In this paper, to sustain a high-quality 24-hour video surveillance (i.e., high reliability) and reduce redundant video traffic volume (i.e., network friendliness), we propose an edge-centric video surveillance system that provides flexible adaptive control of the image enhancement process and video quality based on an event-driven adaptation. In the proposed system, the video bitrate is adaptively controlled according to the contrast of captured videos and conditions in a monitored area (e.g., 'normal', 'caution', and 'alert'). To confirm the system performance, we evaluate objective image quality, accuracy of human detection and video traffic volume generated by the proposed system. Evaluations conclude that the system can reduce the video traffic while sustaining high-quality visibility.

本文言語English
ホスト出版物のタイトル2018 IEEE International Conference on Pervasive Computing and Communications Workshops, PerCom Workshops 2018
出版社Institute of Electrical and Electronics Engineers Inc.
ページ651-656
ページ数6
ISBN(電子版)9781538632277
DOI
出版ステータスPublished - 2018 10月 2
イベント2018 IEEE International Conference on Pervasive Computing and Communications Workshops, PerCom Workshops 2018 - Athens, Greece
継続期間: 2018 3月 192018 3月 23

出版物シリーズ

名前2018 IEEE International Conference on Pervasive Computing and Communications Workshops, PerCom Workshops 2018

Other

Other2018 IEEE International Conference on Pervasive Computing and Communications Workshops, PerCom Workshops 2018
国/地域Greece
CityAthens
Period18/3/1918/3/23

ASJC Scopus subject areas

  • コンピュータ ネットワークおよび通信
  • コンピュータ サイエンスの応用
  • コンピュータ ビジョンおよびパターン認識

フィンガープリント

「Edge-centric Video Surveillance System Based on Event-driven Rate Adaptation for 24-hour Monitoring」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル