Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{204111,
        author = {Parth Sahebrao Gaikwad and Hemal Bhoge and Shivam Swami and Prof Vinita Kute},
        title = {AI BASED CCTV SUSPICIOUS THREAT DETECTION USING CNN},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {1},
        pages = {966-976},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=204111},
        abstract = {With the rise in crimes and security threats to safety & security in our current environment, it’s more important than ever that we ensure safety & security via modern means of technology. One common means of monitoring our premises is through traditional closed-circuit television (CCTV) surveillance systems; however, these systems depend solely on people continually watching over the cameras for continuous monitoring making them inefficient, error-prone with little or no ability to proactively respond to a threat when it happens. As a means of enhancing CCTV surveillance systems functionality to automatically detect suspicious behaviors as they occur through the use of artificial intelligence (AI), this paper proposes an AI-based solution that relies on utilizing deep learning (DL) and computer vision technology to provide real-time detection of suspicious activity. The proposed AI-based CCTV system utilizes different technologies such as OpenCV (Open-Source Computer Vision) video processing technology, TensorFlow (TensorFlow machine learning library) for building/deploying AI models via DL and Python programming language for developing the proposed CCTV surveillance system. This AI-based CCTV surveillance / threat detection system will use convolutional neural networks (CNN) for the analysis of each video frame captured by the CCTV camera and classify that activity as either ‘normal’ or ‘suspicious’. The functionality of the proposed system will be split into various modules, including motion detection; face recognition; behavior analysis; and alert generation; and by automating portions of CCTV surveillance, it will reduce the amount of human intervention needed and provide for a greater accuracy for detection. The ability to generate real-time alerts will enable a quicker response time to a potential threat from time of detection to time of response, and therefore may be deployed in many public locations, organizations, or smart city applications.},
        keywords = {Artificial Intelligence, Closed Circuit Television (CCTV) Surveillance Systems, Deep Learning (DL), Convolutional Neural Networks (CNN), Suspicious Activity Detection, Computer Vision, Real Time Monitoring.},
        month = {June},
        }