A Review of Intelligent Security Research: Progress in Integrated Edge-Cloud-Graph Approaches

DOI: https://doi.org/10.62517/jike.202604106

Author(s)

Yunwei Wang

Affiliation(s)

Shenzhen High School of Mathematics and Physics Shenzhen, Guangdong, China *Corresponding Author

Abstract

This paper provides a systematic review of the current development status and future trends in intelligent security systems. Against the backdrop of accelerating urbanization and growing public safety demands, traditional security systems struggle to meet requirements for real-time responsiveness, intelligence, and robustness. With the rapid advancement of artificial intelligence, edge computing, and the Internet of Things, security systems are progressively transitioning from "passive response" to "proactive prevention and control". This paper focuses on five key research areas: video surveillance and object detection, behavioral analysis and anomaly detection, facial recognition and identity verification, edge computing and resource optimization, and multimodal fusion with security knowledge graphs. Drawing on existing research, it identifies shortcomings in current methods regarding adaptability to complex environments, cross-modal semantic understanding, and edge-cloud collaborative scheduling [1-2]. To address these challenges, this paper proposes constructing an "integrated edge-cloud-graph" intelligent security framework, aiming to upgrade capabilities from mere "visibility" to "comprehension and rapid response". Finally, future research directions are explored, including lightweight edge deployment, unsupervised anomaly detection, multimodal semantic modeling, and dynamic resource scheduling.

Keywords

Intelligent Security; Edge Computing; Multimodal Fusion; Knowledge Graph; Anomaly Detection

References

[1] Wang, F., Jin, X., Zhu, D., Hu, Z., Tang, S., & Lei, J. (2022). Infrared polarised face recognition based on RGB colour space. Laser & Optoelectronics Progress, 59(12), 1233001. [2] Wei, Q., Li, Y., Lou, P., Yan, J., & Hu, J. (2021). Face recognition method based on edge-cloud collaboration. Computer Science, 49(5), 27–47. [3] Zhu, Y., Lai, Z., Li, F., & Huang, Y. (2021). Laboratory attendance system based on face recognition. Journal of Jiangxi University of Science and Technology, 42(3), 115–120. [4] Jiang, S., Yao, K., Chen, L., Wang, Z., & Guo, F. (2023). Masked face recognition method based on CNN–Transformer hybrid model. Journal of Jiangsu University of Technology, 39(2), 55–62. [5] Wu, Y., & Sang, N. (2023). Unsupervised domain adaptation for person re-identification based on consistency constraint and label optimisation. Journal of Huazhong University of Science and Technology (Natural Science Edition), 51(4), 70–78. [6] Chen, R., Peng, Y., & Li, Z. (2023). Water-surface floating object detection method based on continuous unsupervised domain adaptation strategy. Journal of Dalian University of Technology (Hydraulic Engineering Edition), 63(5), 98–105. [7] Wang, Y., Sun, G., Zhao, G., & Xue, H. (2024). Unsupervised novel domain adaptation learning based on self-supervised knowledge. Journal of Nanjing University of Posts and Telecommunications (Natural Science Edition), 44(2), 120–129. [8] Wu, D., Zhang, Z., & Li, Q. (2024). Light and weak target detection method combining machine vision and unsupervised domain adaptation. Journal of Heilongjiang University of Science & Technology, 34(2), 329–336.