Urban parking inefficiency has become a critical challenge in modern cities, leading to increased traffic congestion, higher fuel consumption, and greater environmental impact. This paper proposes an intelligent smart parking management system that integrates hardware sensing, machine learning, and computer vision to enable real-time parking monitoring and automated vehicle identification. The system combines infrared sensors, camera modules, and microcontroller-based control with vision-based parking space detection and automatic license plate recognition (ALPR). Experimental results demonstrate that the parking space detection module achieves an accuracy of 93.97%, while the license plate recognition module attains 84.93% accuracy. Extensive testing under real-world conditions confirms the system's reliability and practicality. The proposed approach enhances parking space utilization, reduces parking search time, and offers a scalable, cost-effective foundation for future smart city parking infrastructure.

This study investigates police officers’ perceptions of advanced technological tools in crime investigations within Abu Dhabi, United Arab Emirates. Drawing on a sample of 375 officers selected through simple random sampling, the research examines five core technologies: crime mapping, facial recognition, car-mounted cameras, body-worn cameras, and license plate readers. Descriptive statistical analysis was employed to evaluate perceptions across dimensions of effectiveness, challenges, and impact on investigative outcomes. The findings reveal generally positive attitudes toward technology adoption, with facial recognition, body-worn cameras, and car-mounted cameras perceived most favourably for their contributions to accountability, transparency, and investigative accuracy. Crime mapping and license plate readers were also valued, particularly for hotspot identification and traffic enforcement, though respondents highlighted challenges related to data accuracy, training adequacy, false positives, and privacy concerns. Overall, the results underscore the importance of organizational capacity, technical support, and ethical safeguards in ensuring that technological innovation enhances policing effectiveness. The study contributes to the discourse on law enforcement modernization in the UAE by providing empirical evidence that can inform policy, guide resource allocation, and strengthen the integration of technology into investigative practice.

Abstract - Road traffic accidents involving two-wheelers remain one of the leading causes of fatalities and serious injuries, particularly in developing countries. A major contributing factor to these accidents is the non-compliance with helmet usage laws. Although traffic regulations mandate helmet use, manual enforcement mechanisms suffer from limitations such as human error, restricted coverage, and high operational costs. To address these challenges, this thesis presents a fully automated Helmet Detection and Fine Generation System using YOLO (You Only Look Once) for object detection and Optical Character Recognition (OCR) for vehicle identification. The proposed system integrates deep learning–based computer vision, text recognition, database management, and automated notification mechanisms into a unified pipeline. The system captures video input from traffic surveillance cameras, detects two-wheeler riders, determines helmet compliance, extracts vehicle number plates using YOLO-based plate detection combined with PaddleOCR, and automatically generates fines for violators. Fine details are securely stored in an SQLite database, and automated email notifications are sent to registered vehicle owners.The system is designed for real-time or near-real-time operation, ensuring scalability, efficiency, and reliability. By reducing dependency on human intervention, the proposed solution enhances enforcement accuracy, promotes road safety, and aligns with smart city initiatives. This thesis details the system architecture, methodology, implementation, and technical specifications, demonstrating the effectiveness of combining YOLO and OCR for intelligent traffic law enforcement. Key Words: Helmet Detection, Two-Wheeler Safety, Traffic Law Enforcement, YOLO, Object Detection, Optical Character Recognition (OCR), PaddleOCR, Number Plate Recognition, Automated Fine Generation, Intelligent Transportation System, Computer Vision, Deep Learning, Smart City, Road Safety, Surveillance Systems, Real-Time Detection, SQLite Database, Automated Notification System

Intelligent transportation systems (ITSs) are crucial for modern traffic management and law enforcement. This paper addresses the challenge of monitoring and managing extensive vehicle traffic in large cities like Lahore, Pakistan. We propose a deep learning based ITS utilizing Vision Transformers combined with convolutional feature extraction to accurately identify vehicle type, color, make/model, and license plates. Experiments were conducted on a comprehensive dataset collected from multiple checkpoints across Lahore under varying environmental conditions. Our proposed model achieved high accuracy rates: 98.0% for vehicle type classification, 96.0% for color detection, 95.0% for make/model identification, and 89.0% for license plate recognition. These results demonstrate the system’s potential to significantly enhance traffic management and road safety and support law enforcement operations in developing urban environments.

A galvanometer-based laser 3D projection system requires accurate mapping between galvanometer control signals and workpiece coordinates to ensure reliable on-part marking. This study presents a calibration and verification pipeline that uses a color camera and a depth sensor to reconstruct 3D target points and estimate the extrinsic parameters between the projector and the workpiece. Laser spot centers are localized in color images, and corresponding depth values are acquired after color–depth alignment. The resulting 3D points are back-projected and transformed into the workpiece coordinate frame. A hybrid solver is employed: the Whale Optimization Algorithm (WOA) provides a global initial estimate, followed by Levenberg–Marquardt (LM) refinement to enhance convergence stability under noisy and small-sample conditions. Experimental validation on an independent 13-point set demonstrates sub-millimeter accuracy, with a mean error of approximately 0.37 mm and a maximum error of 0.87 mm. A further rectangular contour projection test confirms consistent performance, yielding a mean error of 0.434 mm and a maximum error of 0.879 mm, with all errors remaining below 1 mm.

The Hybrid Traffic Safety System is an intelligent, AI-driven traffic monitoring and violation detection platform designed to improve road safety and automate traffic rule enforcement. The system integrates multiple computer vision–based detection modules Automatic Number Plate Recognition (ANPR), Helmet Detection, and Triple Ride Detection within a unified web-enabled architecture. Built using the MERN stack (MongoDB, Express.js, React.js, and Node.js), the platform supports real-time processing, scalable data management, and interactive visualization. AI models developed using TensorFlow, PyTorch, and OpenCV analyze live and recorded surveillance footage to identify vehicles, recognize license plates, and detect rider safety violations with high accuracy. These machine learning components operate as independent microservices and communicate with the backend through secure RESTful APIs or WebSocket connections, enabling efficient separation of computation-intensive tasks from web services. Detected violations are stored along with timestamps, images, and metadata in a centralized database, allowing reliable evidence management and historical analysis. The proposed hybrid architecture enhances system modularity, performance, and extensibility, making it suitable for large-scale urban deployment and smart city environments. By reducing dependence on manual monitoring and enabling continuous, real-time enforcement, the system provides a practical foundation for next-generation intelligent transportation systems aimed at improving traffic compliance and public safety.

The identification of the number plate presents considerable challenges, due to the diversity of types of number plates, which are subject to the environmental conditions. The automated traffic monitoring requires high precision and real-time data processing. This study proposes integrating the YOLOv8 Model with a CNN for Vehicle License Plate Recognition (VLPR). The aim of this study is to construct a VLPR system by using YOLOv8 for real-time vehicle license plate identification and a CNN for character recognition. This research focuses on the vehicle license plate detection and recognition in diverse scenarios. License plates are detected in real time with YOLOv8, even with environmental challenges and partial occlusions of the plates. Performance optimization through real-time adaptive model learning, in which the model dynamically updates based on contextual data, could substantially improve efficiency in numerous practical settings. Expanding the application to include pedestrian and alternative-vehicle identification could provide a basis for a comprehensive traffic monitoring system. A fine-tuned CNN is used to recognize characters on the detected plates. The experiments used a dataset of more than 33,000 license plate images for training the model. The results indicate the system's exceptional reliability and efficiency, rendering it suitable for real-time applications. The experimental results demonstrate that the VLPR system achieves a detection accuracy of 99.4% and a character recognition accuracy of 99.05%.

Laser beams are excellent projection sources due to their high brightness and color purity; however, their high coherence produces speckle noise, which reduces the clarity of images cast by compact projection systems. Existing suppression methods often require complex designs. Here, we propose a simple miniaturized speckle suppression structure (SSS) that consists of a low-absorption particle surface and a micro-vibrating unit. By generating and superimposing different speckle patterns over time, the structure simultaneously reduces both temporal and spatial coherence. A time-varying functional model was developed using a simulation to optimize its dynamic operation. The results of the experimental validation show that at 50 Hz vibration, the speckle contrast decreases from 30.23% to 6.98%, closely matching the simulated prediction of 7.12% and outperforming static configurations by 24%. The results indicate that the SSS is a straightforward, effective solution for enhancing the image quality of compact laser projection displays.

Abstract - The rapid growth of vehicular traffic and increasing violation rates have created a critical need for automated and intelligent traffic monitoring systems. This work presents a Hybrid Traffic Safety System designed to enhance road safety through real-time detection and analysis of traffic violations using artificial intelligence and modern web technologies. The primary aim of the system is to automatically identify common traffic violations such as riding without a helmet, triple riding on two-wheelers, and vehicle identification through Automatic Number Plate Recognition (ANPR). The proposed system employs deep learning and computer vision models developed using TensorFlow, PyTorch, and OpenCV to analyze live and recorded traffic surveillance footage. These AI models operate as independent Python-based microservices and communicate with a MERN stack (MongoDB, Express.js, React.js, Node.js) backend through secure REST APIs and WebSocket connections. Detected violations, along with time-stamped image evidence and metadata, are securely stored in MongoDB and cloud-based storage. A real-time React-based dashboard provides visualization, monitoring, and analytical insights for traffic authorities. Experimental evaluation demonstrates high detection accuracy, low response time, and reliable system performance across varying traffic and lighting conditions. The hybrid architecture improves scalability, modularity, and maintainability while significantly reducing manual monitoring efforts. The proposed system offers a practical and extensible solution for intelligent transportation systems and serves as a strong foundation for smart city traffic management and automated enforcement applications. Key Words: Traffic Safety, Intelligent Transportation System, Artificial Intelligence, Computer Vision, MERN Stack, ANPR, Helmet Detection, Triple Ride Detection.

Academic institutions are not exempt from challenges in managing vehicle transportation, including traffic congestion, inefficient ingress and egress, and the resulting environmental concerns. With the rising volume of vehicles entering and exiting campuses, there is a pressing need for systems that can streamline traffic management while addressing the institution's environmental impact. This problem calls for innovative solutions that balance operational efficiency with sustainability. To address these emerging problems, IdentiPlate has been developed, an integrated system combining automatic license plate recognition (ALPR) with vehicle carbon emission calculations. The system consists of an Android app that utilizes an API for text and image recognition to automatically identify license plates and calculate carbon emissions based on variables such as emission factors, travel distances, and fuel economy data. Additionally, a web-based platform facilitates data management and tracks vehicle ingress and egress. Overall, the IdentiPlate demonstrated its ability to improve operational efficiency and promote environmental responsibility. User feedback indicates that the system was tailored to meet institutional needs, with enhancements suggested to optimize it further.

We present a region-aware, end-to-end motorcycle violation detection pipeline tailored to traffic conditions in Punjab, Pakistan, which integrates three YOLOv11-based components into a unified framework: motorcycle violation detection (MCVD) for helmet compliance and multi-rider analysis, license plate detection (LPD), and license plate character detection (LPCD). The system integrates lightweight object detection, BoT-SORT-based tracking, and character-level recognition, supported by a synthetic-toreal adaptation strategy that combines large-scale synthetic data with limited real samples. Two specific datasets are published, a 40,000-sample synthetic Punjab license plate dataset (PS-LPCD) and a 650-sample real-world dataset (PR-LPCD), which are publicly released in order to encourage research development and adaptation to the region. Class consolidation enhanced MCVD performance (weighted average F1 score: 0.77) and the LPD model performed at mAP50 = 0.99. Two-stage fine-tuning on synthetic and real samples allowed LPCD to reach a character accuracy of ≈ 98% and a full-plate recognition rate of ≈ 90.7%, both surpassing EasyOCR and PaddleOCR, while also achieving lower per-plate latency. With a single motorcycle per frame, the sequential pipeline maintains a throughput of ≈ 9.5 FPS; the throughput reduces in scenes where there are many motorcycles. These findings indicate that synthetic pretraining, together with a small real fine-tuning, can be used to obtain a powerful, scalable, and region aware automatic license plate recognition (ALPR) system, which provides a reproducible method for detecting traffic violations across a variety of license-plate formats.

Automated License Plate Recognition (LPR) systems have become essential for efficient traffic monitoring, parking management, and security applications. Manual methods of license plate identification are time-consuming and prone to errors, especially in high-traffic environments. This paper presents a real-time license plate recognition system using computer vision and optical character recognition techniques. The proposed system captures live video streams through a webcam, detects license plate regions using OpenCV, and extracts alphanumeric characters using EasyOCR. A Flask-based web interface is used to display real-time results, while a MySQL database stores recognized plate numbers along with timestamps and corresponding images. Experimental results demonstrate that the system achieves reliable recognition accuracy with low processing delay under normal lighting conditions. The proposed solution is cost-effective, user-friendly, and suitable for real-time surveillance applications such as traffic control, parking automation, and access management.

A B S T R A C T This project develops a Smart Gate and Parking Management System that automates vehicle access control and real-time parking monitoring using IoT, AI, and cloud computing. The system uses RFID authentication and AI-based number plate recognition (ANPR) to verify vehicles, infrared sensors to detect parking slot occupancy, and a servo motor to control gate access automatically. Data is synchronized with Firebase and displayed on a Django web dashboard for real-time monitoring. The dual-microcontroller architecture (Arduino UNO + ESP8266 NodeMCU) ensures reliable hardware control and seamless cloud communication. The system achieves 93%+ plate detection accuracy, operates with minimal latency, and eliminates manual parking management while supporting future scalability for multiple gates and expanded facilities. Keywords: Smart Parking Management System Automatic Number Plate Recognition(ANPR) YOLOv8 Paddle OCR Machine Learning CNN (Convolutional Neural Network) Internet of Things (IoT) Firebase Realtime Database Django Framework RFID Authentication OpenCV Embedded Systems Python

License plate recognition technology is widely applied traffic management, parking monitoring, and electronic toll collection, among other fields. However, in complex scenarios, such as bright light, fog, rain, snow, and nighttime, there is an urgent need to improve the accuracy of license plate recognition and system robustness. To cope with the difficult problem of license plate recognition in complex scenarios, this study proposes a license plate recognition method based on CSCM-YOLOv8 and CSM-LPRNet. The CPA-Enhancer preprocessing module is used to optimize the input feature representation, and the upsampling quality is improved by the perceptual feature reorganization capability of the CARAFE upsampling module. The SEAM is embedded for adaptive weight allocation, thus enhancing the capability to extract key features. The SEAM is combined with the lightweight C2fMLLABlock convolution module to efficiently aggregate features, thereby maintaining the feature representation capability while reducing the computational cost. The experimental results show that on the dataset used in this study, the CSCM-YOLOv8 network achieves 98.9% accuracy in license plate detection, whereas mAP@0.50-0.95 reaches 58.0%. Compared with the original YOLOv8, the accuracy and mAP@0.50-0.95 are improved by 3.1% and 3.9%, respectively. Moreover, CSM-LPRNet achieves a recognition accuracy of 98.56% in character recognition, which is a 7.0% improvement over that of the original LPRNet. The remarkable performance of this method in complex environments provides an efficient and reliable solution for license plate recognition in intelligent transportation systems.

Vehicle gate access, in general, still relies heavily on manual inspection of identification cards and visual verification by security guards, which is slow, tedious, and susceptible to spoofing. Single-modality, computerized systems that utilize license plates, vehicle appearance, and facial recognition can partially alleviate this difficulty. Still, they are prone to spoofing and generally perform poorly in real-world scenarios (e.g., glare, occlusion, and tinted glass). This study presents TRI-GATE, a tri-modal anti-spoofing framework that unifies vehicle, license plate, and face recognition within a single, real-time decision pipeline. The system employs YOLOv4-tiny for vehicle detection and a MobileNetV2-based classifier for make–model recognition, a retrained MTCNN and LPRNet pair for license plate detection and recognition on Saudi-specific datasets (17,000 images for detection and 35,000 for recognition), and RetinaFace with InsightFace embeddings, along with a linear SVM, for driver identification. An IoU-based best-frame selection scheme reduces latency by forwarding only the most informative frame to the recognition modules. Score-level fusion is then performed by a linear SVM that learns the relative importance of each modality for the final access decision. Evaluated on a dedicated tri-modal dataset, TRI-GATE achieves 97% gate-level accuracy with an end-to-end latency of 66 ms per frame (≈ 15.15 FPS), and demonstrates robust performance in a real-world gate-like deployment, substantially improving both security and operational efficiency over existing single- and bi-modal solutions.

YOLO-B2DV: An Integrated Real-Time Framework for Detection, Tracking and Recognition of Catenary Pole Number Plates in Railway Video Streams

The automation of toll collection is a key problem in modern transportation systems due to increasing traffic congestion, fuel wastage, and dependency on manual or booth-based toll processing. Although FASTag systems improve speed and transparency, they still require physical scanning units and face failures due to unreadable tags or poor sensor contact. To overcome these challenges, this project proposes a Smart Boothless Toll Collection System that operates using RFID-based vehicle identification, Automatic Number Plate Recognition (ANPR) using YOLOv11 and EasyOCR, and energy harvesting through piezoelectric transducers embedded beneath a speed breaker. The system triggers toll processing automatically when a vehicle is detected using IR sensors, captures the number plate using a webcam, and processes it through ANPR for toll deduction. A Flask-based web dashboard logs and displays vehicle history, timestamps, and energy generated, while an RFID module serves as a backup validation mechanism. The result is an automated, energy-aware tolling solution designed for realtime use in Smart Highway infrastructure.

Road safetyfor two-wheeler riders remains a significant challenge, particularly since many people continue to ride without helmets. Manual monitoring of such violations is challenging for traffic authorities because it requires constant attention and a large amount of manpower. To address this problem, this research presents an automated system that combines helmet detection with automatic license plate recognition using modern machine-learning techniques. The system uses deeplearning models, such as YOLO, to identify whether a rider is wearing a helmet by analysing images or video frames in real time. If a rider is found without a helmet, the system automatically detects the motorcycle’s license plate and reads the characters using an OCR-based approach. The proposed method reduces the need for human supervision and increases the speed and accuracy of violation detection. The system is designed to work with CCTV footage and can handle various real-world challenges such as different lighting conditions, camera angles, and backgrounds. By integrating helmet detection and license plate recognition into a single pipeline, this research demonstrates a practical solution for traffic monitoring and enforcement. The work demonstrates that AI-powered systems can play a crucial role in building safer roads, enhancing rule compliance, and supporting smart-city initiatives.

In this project, we built a complete license plate recognition system that uses a camera to take pictures of vehicle plates. The system then uses YOLOv8 to find the plates and EasyOCR to read the text on them. The results are stored in a database that can be searched, including GPS locations and the time the plate was captured. To handle common errors in recognition, the system uses fuzzy matching, which helps avoid saving the same vehicle twice if it's within a certain time and distance limit. Users can look up license plates by number, date, or location through a web interface made with Flask. This interface also shows the history of detections along with map details and images. Our system achieves 91% accuracy in real-world settings, showing how combining databases and computer vision can create a useful tracking tool. The system is designed in a way that makes it easy to use in different real-world situations, such as managing parking or monitoring security. It balances advanced technology with practical use, and handles challenges like changes in lighting and different plate styles.

Ensuring the accuracy of laser beam guidance is a key condition for the quality of modern museum and multimedia installations. Even minor dynamic deviations caused by vibrations or thermal deformations lead to a loss of projection clarity and a decrease in the immersion effect. The purpose of the article is to develop and test an embedded software system to compensate for dynamic errors in laser projection guidance in real time for museum and multimedia installations. The research methodology is mathematical modelling of dynamic deviations, computer vision algorithms for detecting and tracking laser marks, sensor fusion of data from inertial measurement units (IMUs), as well as modular software implementation on single-board computers running Linux. The work uses system analysis to evaluate existing approaches, experimental testing to verify the performance of algorithms, and comparative tests with classical stabilisation methods. The novelty of the research lies in the creation of an affordable and resource-saving system that combines CMOS sensors, light spot detection algorithms and quaternion integration of IMU data. Such an architecture enables the processing of streaming video at a frequency of approximately 90 frames/s, with low hardware requirements, allowing for the high-quality compensation of errors without the need for expensive opto-mechanical equipment. The conclusion of the research. The article identifies the main problems of dynamic stabilisation of laser projections, analyses modern hardware and software solutions, and develops and tests the author’s built-in correction system. The results obtained confirm that the proposed approach enables the enhancement of accuracy and stability in projections for museum and multimedia applications, offering a combination of cost-effectiveness, scalability, and technical reliability.