Curbside Parking Monitoring With Roadside LiDAR

Abstract

Cities worldwide are striving to find more efficient approaches to address the prevalent parking challenges in urban areas. A key aspect of achieving an optimal parking environment is the collection of curbside parking data, which enables informed decision-making and effective management of on-street parking spaces. This study proposes a solution for curbside parking monitoring and data collection using roadside LiDAR systems. By leveraging laser beam variation detection, this solution can extract essential information about parking usage. Unlike existing solutions, such as imagery or embedded sensor-based monitoring, our solution offers portability and ease of deployment for short-term or long-term curbside parking data collection. Additionally, the LiDAR sensor captures only three-dimensional data and is independent of illumination conditions, ensuring stable operation throughout the day while safeguarding privacy by not capturing imagery. These features align with the requirements of city agencies for parking data collection. The workflow follows a simple trend without the need for complex training, as typically seen in machine learning-based methods, and instead relies on parameter tuning based on real-world environmental factors. To validate the effectiveness of our method, we collected curbside parking data for five days at a midtown traffic junction with eight parking spaces. Manual validation confirmed a 95% match between identified parking events and observed data across different time periods. The study further presents parking statistics based on the identified events, revealing crucial insights about parking usage in the study area.