Violent Activity Detection Using Deep Convolutional Neural Networks in Video Surveillance
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Recently, surveillance systems are globally installed for crime prevention by monitoring both private and public places which generate a massive amount of video data. This setup requires human experts to observe and monitor the ongoing activities continuously. To handle this tedious task, an automatic technique workable in real-time for violent activity detection (VAD) is a big challenge. Thus, this paper proposes a three-phase deep learning assisted framework for real-time VAD. In the first phase, a preprocessing step is applied to obtain the salient motion frames from the video data through frame differencing algorithm. These frames are fed into the second phase where a lightweight convolutional neural network (CNN) extracts the most discriminative features. Finally, a CNN model classifies the activity into violent and non-violent scene. In case of violent activity, the activity is reported to notify the nearest security departments and police stations for the prompt action. The proposed method achieves 96% accuracy on the publicly available dataset and outperforms over state-of-the-art methods.
Abstract
Recently, surveillance systems are globally installed for crime prevention by monitoring both private and public places which generate a massive amount of video data. This setup requires human experts to observe and monitor the ongoing activities continuously. To handle this tedious task, an automatic technique workable in real-time for violent activity detection (VAD) is a big challenge. Thus, this paper proposes a three-phase deep learning assisted framework for real-time VAD. In the first phase, a preprocessing step is applied to obtain the salient motion frames from the video data through frame differencing algorithm. These frames are fed into the second phase where a lightweight convolutional neural network (CNN) extracts the most discriminative features. Finally, a CNN model classifies the activity into violent and non-violent scene. In case of violent activity, the activity is reported to notify the nearest security departments and police stations for the prompt action. The proposed method achieves 96% accuracy on the publicly available dataset and outperforms over state-of-the-art methods.
참고문헌 (20)
- T. Hussain, "Intelligent baby behavior monitoring using embedded vision in IoT for smart healthcare centers", Journal of Artificial Intelligence and Systems, 1, pp.110~124 (2019)
- 튜 유 민 울라, "Violence Recognition using Deep CNN for Smart Surveillance Applications", 한국차세대컴퓨팅학회 논문지, 14(5), pp.53~59 (2018), 한국차세대컴퓨팅학회
- F. U. M. Ullah, "An intelligent system for complex violence pattern analysis and detection", International Journal of Intelligent Systems (2021)
- 김민제, "A Study on Person Re-identification in Multi-view Surveillance Videos", 한국차세대컴퓨팅학회 논문지, 17(1), pp.16~24 (2021), 한국차세대컴퓨팅학회[DOI]
- P. Pradeepkumar, "Resource Efficient Edge Computing Infrastructure for Video Surveillance", IEEE Transactions on Sustainable Computing, pp, pp.1~1 (2021)
- A. Hussain, "Anomaly Based Camera Prioritization in Large Scale Surveillance Networks"
- 노먼 칸, "Efficient Sport Videos Classification via Convolutional Neural Network", 한국차세대컴퓨팅학회 논문지, 16(6), pp.7~16 (2020), 한국차세대컴퓨팅학회[DOI]
- I. S. Gracia, "Fast fight detection", PloS one, 10, pp.e0120448~ (2015)
- T. Hassner, "Violent flows: Real-time detection of violent crowd behavior", 2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, pp.1~6 (2012)
- Y. Gao, "Violence detection using oriented violent flows", Image and vision computing, 48, pp.37~41 (2016)
- F. Perronnin, "Improving the fisher kernel for large-scale image classification", European conference on computer vision, pp.143~156 (2010)
- P. Bilinski, "Human violence recognition and detection in surveillance videos", 2016 13th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS), pp.30~36 (2016)
- J. Yu, "Violent scene detection algorithm based on kernel extreme learning machine and threedimensional histograms of gradient orientation", Multimedia Tools and Applications, 78, pp.8497~8512 (2019)
- A. Ullah, "One-shot learning for surveillance anomaly recognition using siamese 3d cnn", 2020 International Joint Conference on Neural Networks(IJCNN), pp.1~8 (2020)
- Z. Liu, "Robust target recognition and tracking of self-driving cars with radar and camera information fusion under severe weather conditions", IEEE Transactions on Intelligent Transportation Systems (2021)
- F. U. M. Ullah, "Diving Deep into Short-Term Electricity Load Forecasting : Comparative Analysis and a Novel Framework", Mathematics, 9, pp.611~ (2021)
- T. Hussain, "Smart and intelligent energy monitoring systems : A comprehensive literature survey and future research guidelines", International Journal of Energy Research (2021)
- S. U. Khan, "Cover the violence: A novel Deep-Learning-Based approach towards violencedetection in movies", Applied Sciences, 9, pp.4963~ (2019)
- Y. Gao, "Violence detection using Oriented VIolent Flows", Image and Vision Computing, 48-49, pp.37~41 (2016)
- I. Serrano, "Fight recognition in video using hough forests and 2D convolutional neural network", IEEE Transactions on Image Processing, 27, pp.4787~4797 (2018)