Open Access Open Access  Restricted Access Subscription Access

VIOLENCE DETECTION IN REAL TIME FOR SURVEILLANCE

Dayes Joseph, Ajay Chandran M, Akshay K, Aswathy K, Vijina P

Abstract


In an interconnected world, the need for advanced surveillance systems capable of detecting criminal offenses and violence is necessary. This paper aims to create a real-time surveillance system using the latest available technology, which focuses on the detection of criminal actions like human fights from surveillance camera footage. The paper makes use of YOLO (You Only Look Once) model, an advanced ML method, which is known for its speed and accuracy in object and action detection. The system employs alert mechanisms to respond to suspicious activities. It extracts its frames and distinguishes between violence and non-violence using data labelling. The YOLO model is fine- tuned to identify the patterns and characteristics of violent actions. Once the model is integrated with the surveillance system it can raise alerts about detected offenses. Human operators can review alerts and take appropriate actions. As YOLO enables the system to provide real-time alerts, It could analyse the videos live and inform concerned security or authority. This helps in taking immediate action on incidents. Several datasets are publicly available or pre-trained models for object detection and action crime recognition can be used. Thus, it supports a diverse range of data and performs various real time activities.


Full Text:

PDF

References


Davies, A.C. and Velastin, S.A.,” A progress review of intelligent CCTV surveillance systems”., (2005) Proc. IEEE IDAACS, pp.417-423.

Wang, D., Zhang, Z., Wang, W., Wang, L. and Tan, T.,” Baseline results for violence detection in still images”.,(2012), IEEE Ninth International Conference on Advanced Video and Signal-Based Surveillance (pp. 5457). IEEE.

P. Bilinski and F. Bremond, (2016)” Human violence recognition and detection in surveillance videos”,in Proc. 13th IEEE Int. Conf. Adv. Video SignalBased Surveill. (AVSS), Aug.pp. 30–36.

Georgios Georgakis, Arsalan Mousavian, Alexander C. Berg†, Jana Koseck” Synthesizing Training Data for Object Detection in Indoor Scenes”(2017)

Jim Lee, Zifeng Yang, and Honggang Zhan” Efficient Violence Detection Using3D Convolutional Neural Networks” (2019)

R. Vijeikis, V. Raudonis, and G. Dervinis.,” Efficient violence detection in surveillance” (2022)

Swathikiran Sudhakaran and Oswald Lanz.,” Learning to Detect Violent Videos using Convolutional Long Short- Term Memory” (2017).

Fath U Min Ullah, Mohammad S. Obaidat, Amin Ullah, Khan Muhammad, Mohammad Hijji, Sung Wook Baik.,” A Comprehensive Review on Vision-based Violence Detection in Surveillance Videos” (2022)

Batyrkhan Omarov, Sergazi Narynov, Zhandos Zhumanov, Aidana Gumar, and Mariyam Khasanova.,” A Systematic Review of Violence Detection Techniques in Video Surveillance Security Systems” (2022),2010, 43, 378-386

VIKTOR D ENUS HUSZ A R, VAMSI KIRAN ADHIKARLA (Member, IEEE), IMRE N EGYES, AND CSABKRASZNAY” Toward Fast and Accurate Violence Detection for Automated Video Surveillance Applications” (2023)


Refbacks

  • There are currently no refbacks.