Aug 2026· IAES International Journal of Artificial Intelligence (IJ-AI)· Vol 15, pp. 3176· 0 citations· 27 references
TL;DR
This study represents the design and development of an access control system that includes accurate detection of essential PPE items, integration of facial recognition for identity verification, real-time monitoring of video feeds, and an intuitive user interface for security personnel to manage access and compliance efficiently.
Abstract
This paper presents an artificial intelligence (AI)-powered automated access control system that aims to reduce delays and improve safety. The primary problem addressed is effective monitoring of compliance with personal protective equipment (PPE) and secure access control for personnel entering sites. This study represents the design and development of an access control system that includes accurate detection of essential PPE items (e.g., safety helmets, gloves, goggles, and gas detectors), integration of facial recognition for identity verification, real-time monitoring of video feeds, and an intuitive user interface for security personnel to manage access and compliance efficiently. The software part uses you only look once (YOLO) version 8 for real-time object detection, classification, and drawing the bounding boxes around the detected object in a single forward pass. The hardware platform consists of NVIDIA Jetson AGX Orin 64 GB as an edge computing device. The developed AI-based embedded system is tested and validated with real-world scenarios and achieved a mean average precision (mAP) of 98.4% for PPE detection and 99.38% accuracy for face recognition.
The high rate of motor vehicle theft in Bandung City indicates that conventional security surveillance systems still have limitations in providing effective and responsive monitoring. Therefore, this study aims to design and implement a smart Closed Circuit Television (CCTV) face detection system based on the Internet of Things (IoT) and Artificial Intelligence (AI) capable of performing real-time monitoring and face detection. The research employed the User Centered Design (UCD) method, which consists of user needs identification, system design, implementation, and evaluation. The system was developed using a Raspberry Pi as the IoT edge device, a camera for image acquisition, the InsightFace algorithm for face recognition, and Firebase and the Telegram Bot API for data storage and notification delivery. The novelty of this study lies in the development of a smart CCTV system that integrates the User Centered Design (UCD) method, Internet of Things (IoT) technology, the InsightFace algorithm based on Artificial Intelligence (AI), an automatic pan-tilt module, and Telegram notifications into a single real-time security monitoring platform. The results show that the system achieved a face recognition accuracy of 92%, delivered Telegram notifications with an average response time of 2.8 seconds, and performed object tracking using the pan-tilt module with a response time of less than 0.5 seconds. In addition, the system successfully detected unknown faces, synchronized data to Firebase instantly, managed video storage automatically through a rolling buffer mechanism, and provided a responsive web-based dashboard. Based on the overall testing results, the developed system effectively improved security monitoring, facilitated remote surveillance, and met user requirements in accordance with the User Centered Design (UCD) approach.
Modern academic campuses face a growing tension between convenient vehicle access and increasing demands for security and operational efficiency. At scale, manual gate checks are slow, error-prone, and difficult to audit. This paper presents the design, implementation, and evaluation of an intelligent License Plate Recognition (LPR) system for smart-campus access control. The system integrates a fine-tuned YOLOv8 detector for vehicle and plate localization, the EasyOCR engine for real-time plate-text recognition, a MySQL relational database that matches recognized plates against authorized and blacklisted vehicle lists, and a Django web dashboard that provides administrative visibility. A blacklist-aware alerting subsystem dispatches SMS and e-mail notifications whenever an unauthorized or blacklisted vehicle is detected. The novelty of this work lies not in any single algorithmic component but in the complete, reproducible, open-source integration of detection, recognition, database matching, logging, and alerting into one deployable framework adapted to local plate formats and campus operational requirements. The detector was trained on a combination of public benchmarks and a locally collected campus dataset and evaluated on a held-out test set of 850 images, achieving a plate-detection mAP@0.5 of 0.942, precision of 0.945, recall of 0.938, full-plate OCR accuracy of 92.1%, end-to-end recognition accuracy of 94.8%, and an average throughput of 23.5 FPS on commodity hardware. The architecture is modular, scalable, and deployable on Commercial Off-The-Shelf (COTS) hardware, offering an affordable and replicable solution for budget-constrained institutions.
B. Abdulrahman, Mohammed Faraj· Sulaimani Journal for Engine...· 0 citations
In today’s digital world, ensuring secure and reliable authentication has become essential in workplaces and educational institutions. Traditional attendance methods, such as manual registers and card-based systems, often prove to be inefficient, error-prone, and vulnerable to issues like proxy attendance and data manipulation. Although biometric techniques like fingerprint and card-based authentication are widely used, they still face limitations in terms of scalability, speed, and user convenience. Facial recognition has emerged as a contactless and user-friendly biometric solution that provides a more secure and efficient approach to identity verification. Additionally, these systems often lack proper integration of time-based tracking features. This work focuses on the design and implementation of an AI-based facial recognition system for automated employee attendance and time logging. The proposed system utilizes advanced machine learning and computer vision techniques to detect and recognize faces in real time while accurately recording entry and exit times. By incorporating efficient feature extraction and one-shot learning methods, the system is capable of delivering reliable performance even with limited training data. Overall, this approach minimizes manual effort, prevents proxy attendance, and enhances the integrity of attendance records. Keywords— Facial Recognition, Automated Attendance System, Temporal Logging, Machine Learning, Computer Vision, One-Shot Learning, Biometric Authentication, Real-Time Face Detection, Workforce Management, Data Integrity
Chanda Aakash, C. Manikanta, Gorle Sai Chaitanya et al.· International Journal of Eng...· 0 citations
In the context of increasing pressure on healthcare systems, the application of automation and artificial intelligence (AI) has become a key solution for optimizing healthcare delivery processes. This study proposes the design of a smart medicine cabinet system based on a microcontroller platform, such as STM32, integrated with IoT and AI technologies to automate medication dispensing. The system employs a camera and authentication algorithms, including facial recognition or PIN-based verification, in combination with image processing techniques to ensure that medication is delivered to the correct patient while also supporting the assessment of the patient’s current condition. Through appropriate model training, the cabinet is expected to identify the patient’s health status by analyzing visual features extracted from images, particularly skin-related characteristics. With a hardware architecture comprising servo motors for compartment control, an alarm speaker, and a display interface, the smart cabinet not only reminds patients to take their medication but also assists healthcare personnel in remotely managing treatment data via a mobile application. Experimental results indicate that the system operates stably, minimizes medication-related errors, reduces the distance between hospitals and patients, and is particularly beneficial for supporting elderly individuals and preventing cross-infection in hospital environments.
Ho-Si-Hung Nguyen, H. Nguyen, T. Lê et al.· E3S Web of Conferences· 0 citations
The paper explores how the classical approaches to image processing have been transformed to deep learning based methods such as their application in object detection and tracking, activity recognition, anomaly detection and facial recognition, and provides the future research direction, which is important to the next-generation intelligent surveillance systems.
Ajay Krishnan· International Journal of Mod...· 0 citations
A YOLOv8-based visual detection application in ONNX format to identify safety helmet violations in real-time to demonstrate the application's stability and accuracy in computer vision-based automated surveillance.
S. Syufrijal, Heri Firmansyah, Christophorus Mrc Yuda et al.· EPJ Web of Conferences· 0 citations