An IoT-Based Smart Parking Management System with Automatic Number Plate Recognition and Environmental Hazard Monitoring
Abstract
Parking is a growing problem in urbanizing Nepalicities, where many parking lots still rely on manual occupancyrecording, vehicle logging, and cash transactions. Althoughpast IoT applications have looked into parking systems, NepaliANPR, reservation management, and automatic parking paymentshave been investigated separately in earlier research,there are very few cases where all these aspects are combinedtogether along with environmental hazard detection into a single,low-cost physical model.This paper presents a connected IoTbasedsmart parking system using ESP32/Arduino sensors andactuators, a React.js dashboard backed by Firebase RTDB, anda tiered Nepali ANPR pipeline. Plate localization and characterdetection are performed using YOLOv8s, a fine-tuned 34-classCNN, and geometrical reconstruction to assemble individuallydetected characters into a full plate string. The object detectionframework had 95.44% mAP@0.50 (97.92% & 93.09% for Recalland precision, respectively) after held-out tests on a datasetcontaining 867 images. Overall, the ANPR Pipeline had awhole-plate detection accuracy of 53.73% on raw 67 entry/exitpictures without cropping (36/67; 95% CI: 41.9-65.1%), usingfuzzy whole-plate character matching with an 80% threshold.In view of the small number of end-to-end test data, this findingshould be seen more as an initial performance estimate than asa statistically reliable indication of deployment-based accuracyfor ANPR. The occupancy detection component performed at97.14% accuracy from 70 test samples manually inspected, whilethe reservation, transaction log maintenance, and hazard detectioncapabilities were also demonstrated on the real-world systemwithout quantitative evaluation. These outcomes demonstrate thefeasibility of combining localized ANPR technology with low-cost,IoT-enabled parking infrastructure in Nepal, although large-scaleimplementation should await further end-to-end evaluation andfield trials.