AI-powered threat detection in surveillance systems: A real-time data processing framework

Emmanuel Cadet 1, *, Olajide Soji Osundare 2, Harrison Oke Ekpobimi 3, Zein Samira 4 and Yodit Wondaferew Weldegeorgise 5

1 Riot Games, California, USA.
2 Nigeria Inter-Bank Settlement System Plc (NIBSS), Nigeria.
3 Shoprite, Capetown, South Africa.
4 Cisco Systems, Richardson, Texas, USA.
5 Deloitte Consulting LLP, Dallas, TX, USA.
 
Review
Open Access Research Journal of Engineering and Technology, 2024, 07(02), 031–045.
Article DOI: 10.53022/oarjet.2024.7.2.0057
Publication history: 
Received on 01 September 2024; revised on 11 October 2024; accepted on 14 October 2024
 
Abstract: 
The increasing need for enhanced security has driven the adoption of AI-powered threat detection in surveillance systems. Traditional surveillance methods, reliant on manual monitoring, are often inefficient in detecting complex, evolving threats in real time. This review proposes a comprehensive real-time data processing framework for AI-powered threat detection in surveillance systems, designed to automate and optimize threat identification, classification, and response. The framework integrates AI algorithms, including machine learning and deep learning models, to analyze vast amounts of surveillance data from various sources such as video feeds, audio recordings, and sensor inputs. It utilizes techniques like object detection, facial recognition, and anomaly detection to identify potential threats, while leveraging stream processing frameworks (e.g., Apache Kafka, Apache Flink) to ensure low-latency, real-time analysis. Edge computing is incorporated to reduce network bottlenecks and enable faster decision-making closer to the data source. The framework also addresses the challenges of high data volume and velocity, as well as the need for scalable, flexible infrastructure. Security measures such as encryption, identity and access management (IAM), and compliance with data privacy regulations ensure that sensitive information is protected. The inclusion of continuous model training allows the system to adapt to emerging threats and reduce false positives and negatives. Case studies from urban environments, critical infrastructure, and law enforcement demonstrate the practical applications and effectiveness of this AI-driven approach. By integrating real-time data processing with advanced AI models, the framework provides a robust solution for improving the efficiency and accuracy of threat detection in modern surveillance systems. This research contributes to the growing field of AI-enhanced security, paving the way for future advancements in intelligent surveillance.

 

Keywords: 
Artificial intelligence; Threat Detection; Surveillance Systems; Review
 
Full text article in PDF: