Smart Parking System Over Distributed Node Network using Computer Vision

Abstract

The growing demand for urban parking has led to significant challenges, particularly with conventional centralized smart parking systems that often face single points of failure, limited scalability, and complex deployment processes. Existing solutions are inadequate in addressing these issues, prompting the need for a more resilient approach. This research presents a distributed, cost-effective, innovative parking system to overcome these limitations. The proposed system features a network of nodes strategically deployed across parking lots and interconnected through wireless LAN (Local Area Network), enabling real-time data access through a cloud-based website and back-end server. By utilizing low-power cameras for object detection, the system enhances operational efficiency and accuracy while minimizing costs. Additionally, the Google Maps API (Application Programming Interface) integration provides a user-friendly interface, facilitating easy navigation to available parking spaces. This solution explicitly targets the unique parking challenges in India, demonstrating promising applications for parallel parking scenarios.