Tehnički vjesnik, Vol. 33 No. 6, 2026.
Pregledni rad
https://doi.org/10.17559/TV-20251214003202
Video Anomaly Detection for Smart Lecture Halls: A Systematic Review and Engineering Framework for Privacy-Aware Educational Monitoring
Bing Liu
; Chengdu Aeronautic Polytechnic University, No. 699 East 7th Road, Checheng, Longquanyi District, Chengdu, China
Dong Zhao
; Chengdu Aeronautic Polytechnic University, No. 699 East 7th Road, Checheng, Longquanyi District, Chengdu, China
Renjun Huang
; Chengdu Aeronautic Polytechnic University, No. 699 East 7th Road, Checheng, Longquanyi District, Chengdu, China
Fang Yang
; Chengdu Aeronautic Polytechnic University, No. 699 East 7th Road, Checheng, Longquanyi District, Chengdu, China
Chun Sheng Li
; Geely University of China, No. 123, Section 2, Chengjian Avenue, Eastern New District, Chengdu, China
*
* Dopisni autor.
Sažetak
Video anomaly detection (VAD) has been extensively studied in urban surveillance and industrial safety; however, its application to educational environments such as university lecture halls remains limited. Lecture halls pose distinct engineering challenges, including dense crowds, long-duration activities, subtle behavioral anomalies, and strict privacy requirements. This paper presents a systematic review of VAD methods published between 2015 and 2025, focusing on their applicability to smart lecture hall scenarios. Following PRISMA guidelines, 57 relevant studies were selected and analyzed. Based on this analysis, the paper provides three technical contributions: (i) a domain-specific taxonomy of VAD approaches evaluated against lecture-hall constraints, (ii) a reference multimedia system architecture that integrates edge computing and privacy-preserving processing, and (iii) a dataset blueprint defining standardized protocols for privacy-compliant lecture-hall anomaly data collection. By organizing existing methods within an engineering-oriented framework, the study offers practical guidance for researchers and practitioners developing intelligent, privacy-aware educational monitoring systems.
Ključne riječi
dataset blueprint; deep learning; edge computing; privacy preservation; smart lecture halls; system architecture; systematic review; video anomaly detection
Hrčak ID:
353104
URI
Datum izdavanja:
7.11.2026.
Posjeta: 0 *