Computer Vision for Weapon Detection in Educational Environments: A Systematic Literature Review

Abstract

This study presents a systematic review of the literature on the use of computer vision algorithms for weapon detection in educational environments. Through the analysis of 13 selected studies from an initial corpus of 10,519 articles, the results demonstrate that models based on Convolutional Neural Networks, particularly variants of YOLO, are predominantly used due to their high accuracy and real-time efficiency. This work highlights the need for technological advancements to address challenges such as the variability of weapon types and the diverse school scenarios. Furthermore, the practical implications of these technologies in enhancing school security and the importance of ethical and privacy considerations are discussed. The review also reveals significant gaps in current research, such as the lack of studies focused on specific educational environments and the need for more representative and diverse datasets.