Title: An Intelligent Computer Vision-Based Campus Access Control and Monitoring System: A Case Study of Ruaha Catholic University (RUCU)
Authors: Eziberi Chuma¹, Paul Kitangita², Abdulkadir Mohammed³, Christina Ganagwa?, Faston Kataita?, Hamisi Mihezo?, Dayness Urio? and Danny Mfungo?
Volume: 10
Issue: 4
Pages: 1-9
Publication Date: 2026/04/28
Abstract:
Campus security breaches pose significant risks to student safety and institutional assets, particularly through unauthorized access via fake or borrowed identification cards. Traditional manual verification methods visual inspection of ID cards and logbook sign-ins-are time-consuming, subjective, prone to human error, and difficult to scale across multiple entry points. This leads to delayed response to security incidents, overcrowded entry gates during peak hours, and lack of actionable data for security management. Advances in computer vision and deep learning now enable automated facial recognition systems for rapid and accurate identity verification. Cameras capture faces at entry points, and lightweight models provide instant authentication results (identity match, confidence score, access decision) either on edge devices or via campus servers. This review examines computer vision approaches for campus access control, including key architectures (LBPH, CNN-based face recognition), datasets, and deployment strategies. It highlights high accuracy in controlled laboratory conditions but significant performance degradation in real campus environments due to variable lighting, occlusions (masks, glasses), pose variations, and demographic bias. The proposed Intelligent Campus Access Control and Monitoring System is evaluated against existing implementations, showing good alignment with institutional security needs while identifying gaps in privacy protection, real-time scalability, and multi-modal verification. Recommendations focus on real-world data collection across diverse conditions, bias mitigation, edge computing for privacy-preserving inference, integration with existing campus management systems, and user acceptance testing with diverse campus populations.