Intelligent video surveillance has evolved rapidly over the past two decades, driven by advances in computer vision, artificial intelligence, and edge computing. A new systematic review published in the Journal of Big Data takes an unusually broad look at the field, arguing that most existing literature reviews examine isolated tasks such as object detection or motion tracking without addressing how these components integrate within a complete surveillance system.
The review, led by Mahmoud Elnady and colleagues, covers research published between January 2000 and July 2025. The team searched major digital repositories, including IEEE Xplore, ScienceDirect, Google Scholar, arXiv, Springer, and the ACM Digital Library, to assemble a system-level picture of how intelligent surveillance architectures and methodologies have developed over roughly twenty-five years.
The authors identify three main contributions. First, they propose a structured taxonomy of intelligent surveillance systems organized around key functionalities: object detection, motion tracking, abnormal event detection, vehicle and traffic monitoring, object counting, and multi-camera coordination.
Second, the paper offers a comparative analysis of the methodologies, datasets, and platforms used across the studies it surveys, placing these elements side by side across a body of literature that has largely been examined task by task.
Third, the authors explore the interconnections among the core tasks, examining how the individual functionalities relate to one another within a complete system rather than as standalone algorithms.
Among the critical challenges the review identifies are real-time processing, privacy preservation, data scalability, and system interoperability, all of which shape how surveillance systems are designed and connected across hardware and software platforms.
The review also flags underexplored research areas, naming federated learning and cross-camera integration in particular as directions that have received comparatively little attention in the literature it analyzed.
As a practical output, the authors distill guidelines for designing next-generation surveillance systems, aimed at helping developers address the integration gaps and open problems their synthesis reveals. Because the article is being shared early as a peer-reviewed, accepted manuscript, the published version is subject to further edits and will be replaced by the final Version of Record.
For a field that has expanded alongside progress in computer vision, artificial intelligence, and edge computing, the review’s system-level perspective offers a consolidated map of where intelligent video surveillance research stands and where its most significant technical questions remain open.
Subject of Research: A systematic review of intelligent video surveillance system architectures, methodologies, and research challenges
Article Title: Survey on intelligent video surveillance
Article References: Elnady, M., Abdelmunim, H. E., Abidou, D., & Elmadah, I. (2026). Survey on intelligent video surveillance. Journal of Big Data, 13(1), Article 166. https://doi.org/10.1186/s40537-026-01560-z
Image Credits: AI Generated
DOI: 10.1186/s40537-026-01560-z
Keywords: intelligent video surveillance, computer vision, artificial intelligence, edge computing, object detection, motion tracking, anomaly detection, multi-camera coordination, federated learning, privacy preservation, systematic review, Journal of Big Data
News Source: Blake Davidson. (October 23, 2026). Survey Maps Two Decades of Intelligent Video Surveillance Systems. Scienmag.
