Enhancing Video Surveillance: A Fusion of Spatial Zone Analysis and Temporal View Profiling for Suspicious Activity Detection

カテゴリ: 研究会(論文単位)

論文No: CMN24030

グループ名: 【C】電子・情報・システム部門 通信研究会

発行日: 2024/03/25

タイトル(英語): Enhancing Video Surveillance: A Fusion of Spatial Zone Analysis and Temporal View Profiling for Suspicious Activity Detection

著者名: Mitra Abhishek(Hitachi India Pvt. Ltd),Kamoshida Ryota(Hitachi India Pvt. Ltd),Ganesh Ananth(Hitachi India Pvt. Ltd),Banerjee Anoushka(Hitachi India Pvt. Ltd)

著者名(英語): Abhishek Mitra(Hitachi India Pvt. Ltd),Ryota Kamoshida(Hitachi India Pvt. Ltd),Ananth Ganesh(Hitachi India Pvt. Ltd),Anoushka Banerjee(Hitachi India Pvt. Ltd)

キーワード: Suspicious activity detection|Video surveillance|Zone analysis|view profiling|loitering detection |behavioural pattern|Suspicious activity detection|Video surveillance|Zone analysis|view profiling|loitering detection |behavioural pattern

要約(日本語): This paper presents a novel approach to enhance video surveillance systems by integrating spatial zone analysis and temporal view profiling for improved detection of suspicious activities. The proposed framework aims to address the limitations of traditional surveillance methods by combining spatial and temporal information to achieve a more meaningful suspicious activity detection mechanism. Spatial zone analysis involves partitioning the surveillance area into distinct zones and analysing each person dwell time and movement pattern across different zones. Simultaneously, temporal view profiling captures person’s viewing patterns and behaviours over time, enabling the system to discern abnormal activities based on temporal dynamics. The fusion of these two approaches results in a comprehensive surveillance system capable of detecting and identifying suspicious activities with higher confidence. The implementation of the proposed framework demonstrate the effectiveness in suspicious activity detection.

要約(英語): This paper presents a novel approach to enhance video surveillance systems by integrating spatial zone analysis and temporal view profiling for improved detection of suspicious activities. The proposed framework aims to address the limitations of traditional surveillance methods by combining spatial and temporal information to achieve a more meaningful suspicious activity detection mechanism. Spatial zone analysis involves partitioning the surveillance area into distinct zones and analysing each person dwell time and movement pattern across different zones. Simultaneously, temporal view profiling captures person’s viewing patterns and behaviours over time, enabling the system to discern abnormal activities based on temporal dynamics. The fusion of these two approaches results in a comprehensive surveillance system capable of detecting and identifying suspicious activities with higher confidence. The implementation of the proposed framework demonstrate the effectiveness in suspicious activity detection.

本誌: 2024年3月28日-2024年3月29日通信研究会

本誌掲載ページ: 73-76 p

原稿種別: 英語

PDFファイルサイズ: 1,753 Kバイト