Zhang, Yuxuan
(2024)
Drive-by sensing for on-street parking spot detection.
Master's Thesis, University of Pittsburgh.
(Unpublished)
Abstract
On-street parking remains a persistent challenge, leading to driver frustration and wasted time. More critically, drivers circling streets in search of parking contribute to traffic congestion, increased carbon emissions, and unnecessary fuel consumption. The crux of the on-street parking dilemma lies in the insufficient awareness of available parking spaces. To tackle this issue, various solutions, including static and mobile sensing methods, have been explored and implemented. Yet, these strategies have encountered obstacles that hinder widespread adoption. Static sensors, for instance, are typically limited to monitoring a single parking space each, leading to high costs for comprehensive coverage. Mobile sensing strategies, on the other hand, aim to maximize sensor utility by collecting data on multiple spaces. However, these methods have traditionally required specialized hardware installations on vehicles, posing barriers to large-scale application.
In this thesis, we introduce an innovative passive mobile sensing solution that mitigates the need for dedicated hardware installation. Our key observation is that moving vehicles inherently emit signals, predominantly in the form of tire noise and aerodynamic noise. When parked cars are present along the roadside, these sounds reflect back to the moving vehicle. By leveraging a smartphone to capture these naturally generated signals, we can effectively differentiate between empty spaces and parked cars. To realize this idea, we have developed an end-to-end system that achieves equal performance with the state-of-art mobile sensing technologies in detecting available on-street parking spots. Our system comprises a pre-processing module that employs signal processing techniques for automatic data segmentation, a Deep Neural Network (DNN) model for parking spot availability prediction, and a post-processing component that refines the parking information. To support our model, we created our own dataset from data collected over 2 weeks, including 2,512 samples in total. Our approach not only demonstrates the feasibility of hardware-free, software-based solutions for on-street parking spot detection but also holds the potential for scalable implementation across urban environments.
Share
| Citation/Export: |
|
| Social Networking: |
|
Details
| Item Type: |
University of Pittsburgh ETD
|
| Status: |
Unpublished |
| Creators/Authors: |
|
| ETD Committee: |
|
| Date: |
13 May 2024 |
| Date Type: |
Publication |
| Defense Date: |
26 March 2024 |
| Approval Date: |
13 May 2024 |
| Submission Date: |
2 April 2024 |
| Access Restriction: |
No restriction; Release the ETD for access worldwide immediately. |
| Number of Pages: |
58 |
| Institution: |
University of Pittsburgh |
| Schools and Programs: |
School of Computing and Information > Computer Science |
| Degree: |
MS - Master of Science |
| Thesis Type: |
Master's Thesis |
| Refereed: |
Yes |
| Uncontrolled Keywords: |
passive mobile sensing, machine learning, signal processing, on-street parking spot detection. |
| Date Deposited: |
13 May 2024 17:30 |
| Last Modified: |
13 May 2024 17:30 |
| URI: |
http://d-scholarship.pitt.edu/id/eprint/45968 |
Metrics
Monthly Views for the past 3 years
Plum Analytics
Actions (login required)
 |
View Item |
![[feed]](/images/feed-icon-32x32.png){kind=icon}