Automated Parking Management for Urban Efficiency: A Comprehensive Approach

Tomasz Ludwisiak; Magdalena Mazur-Milecka

doi:10.1109/hsi61632.2024.10613584

Automated Parking Management for Urban Efficiency: A Comprehensive Approach

Abstract

Effective parking management is essential for ad-dressing the challenges of traffic congestion, city logistics, and air pollution in densely populated urban areas. This paper presents an algorithm designed to optimize parking management within city environments. The proposed system leverages deep learning models to accurately detect and classify street elements and events. Various algorithms, including automatic segmentation of urban landscapes, occupancy detection, and illegal parking violation detection, were developed and tested. To validate the system, cameras were installed facing streets and parking areas, and all algorithms were tested on real-world data. The segmentation network achieved a mean Average Precision (mAP) of 0.791. The occupancy detection algorithm showed a precision and recall of 0.97 on parking camera data, while the illegal parking violation detection system achieved precision and recall values of 0.971 and 0.958, respectively. This research contributes to smarter, more efficient urban parking solutions, enhancing overall city management.