AI-Based CCTV Anomaly and Fire Detection System

ThispaperhaspresentedacomprehensiveAI-basedCCTV surveillance system integrating real-time behavioral anomaly detection, fire and smoke recognition, IoT-driven emergency response,andweb-basedmonitoringintoaunified,deployable platform. The system achieves 91.3% overall anomaly detectionaccuracy, 94.7%firedetectionprecision,andend-to- end alert latency under 1.2 seconds—a substantial improvement over both traditional computer vision baselines and commercial alternatives. TheautonomousIoT firesuppression capability, with total actuation latency of approximately 1.37 seconds, represents a meaningfuladvance beyond passive alerting systems, offering potential to significantly reduce property damage and casualties in fire emergencies. The multi-tiered risk scoring frameworkprovidesoperatorswithactionable,graduatedalerts ratherthanbinaryalarms,reducingalertfatigueandsupporting more nuancedincident management. Future work will expand the anomaly taxonomy to include weapon detection, deploy federated learning across distributed camera networks, and investigate edge inference on dedicated hardware accelerators for improved fault tolerance.

[1]   A. Brodsky, “The economics of CCTV monitoring: scale, staffing, and technology,” Secur. J., vol. 34, no. 2, pp. 112–128, 2021. [2] M.TicknerandJ. Bhatt, “Limits ofsustainedattentioninsecurity monitoring tasks,” Hum. Factors, vol. 63, no. 4, pp. 601–613, 2021. [3] Y.LeCun, Y. Bengio, and G. Hinton, “Deep learning,” Nature, vol. 521, no. 7553, pp. 436–444, 2015. [4]   A. Karpathy et al., “Large-scale video classification with convolutional neuralnetworks,”inProc. IEEE CVPR, Columbus, OH, 2014, pp. 1725–1732. [5] R.Cucchiara, C. Grana, M. Piccardi, and A. Prati, “Detecting moving objects, ghosts, and shadows in video streams,” IEEE Trans. PatternAnal.Mach. Intell., vol.25, no. 10, pp.1337–1342, 2003. [6] M.Hasan, J. Choi,J. Neumann, A. K. Roy-Chowdhury, andL.S. Davis, “Learning temporal regularity in video sequences,” inProc. IEEE CVPR, Las Vegas, NV, 2016, pp. 733–742. [7] W.Luo, W. Liu, and S. Gao, “A revisit of sparse coding based anomaly detection in stacked RNN framework,” in Proc. IEEE ICCV, Venice, Italy, 2017, pp. 341–349. [8]   W. Sultani, C. Chen, and M. Shah, “Real-world anomaly detection in surveillance videos,”in Proc. IEEE CVPR, Salt Lake City, UT, 2018, pp. 6479–6488. [9] T.-H.Chen, P.-H. Wu, and Y.-C. Chiou, “An earlyfire-detection method based on image processing,” in Proc. IEEE ICIP, Singapore, 2004, pp. 1707–1710. [10]  Z. A. Khan, W. Jawad, and A. Usman, “Fire detection in still images using convolutional neural networks,” in Proc. IEEESMC, Bari, Italy, 2019, pp. 1018–1023. [11]   D. Wagner and P. Nguyen, “Automated emergency response systems:a surveyof commercial IoT solutions,”IEEE IoT J., vol. 9, no. 7, pp. 5201–5215, 2022. [12]   Z. Zivkovic, “Improved adaptive Gaussian mixture model for background subtraction,” in Proc. IEEE ICPR, Cambridge, UK, 2004, pp. 28–31. [13]   E. Ince, M. Ozalp, and I. Basyigit, “AIDER: Aerial image datasetforemergencyresponse,”IEEEAccess,vol.9,pp.72422– 72435, 2021.