On-street Parking Localization via Speed-filtered GPS Probe Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current digital maps struggle to accurately identify and update on-street parking spaces due to labor-intensive manual collection and image analysis, which becomes outdated, and existing GPS probe data techniques are inadequate for differentiating on-street parking in dynamic cities.
Innovation Solution
A method that receives probe points from vehicles along a roadway, selects points based on vehicle speed, compares them with a known roadway representation, and determines potential parking spaces using shape algorithms and imaging technology for verification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual collection or image analysis is used to gather parking space information, then the information can be obtained, but the process becomes labor intensive and the information becomes outdated over time
Solution Approach 1:
The patent replaces manual mechanical collection methods with automated GPS probe data processing. Vehicles equipped with GPS sensors automatically generate probe points that are processed by algorithms to identify parking spaces, eliminating labor-intensive manual collection while maintaining high accuracy through continuous data streams from multiple vehicles.
Solution Approach 2:
The system enables parking space information to be self-updated through continuous collection of GPS probe data from vehicles in the area. The automated processing of probe points allows the system to automatically detect and update parking space availability without human intervention, ensuring information remains current.
2Measurement precision
If traditional deterministic and rule-based approaches are used to detect parking spaces from GPS probe data, then off-street parking spaces can be extracted, but the approach performs poorly when probe data becomes large in size and cannot differentiate on-street parking spaces
Solution Approach 1:
The patent applies different processing rules and algorithms tailored to specific parking space characteristics. On-street parking detection uses criteria such as proximity to roadways and specific spatial patterns, while off-street parking uses different criteria appropriate for parking lots. This localized approach allows accurate detection of both parking types despite their fundamental differences.
Solution Approach 2:
The system dynamically adapts its detection algorithms based on the type of parking space being analyzed. Rather than using fixed deterministic rules, the system adjusts its processing logic according to whether it is detecting on-street or off-street parking, enabling versatile application across different parking environments while maintaining high accuracy.
3Productivity
If GPS probe data is used to identify parking spaces, then automated detection is possible, but it becomes difficult to differentiate parking spaces from noise when probe data is large in size
Solution Approach 1:
The patent introduces intermediate processing steps including probe point filtering, clustering algorithms, and validation rules that act as mediators between raw GPS probe data and final parking space identification. These intermediate layers filter out noise and false positives while preserving genuine parking space signals, enabling accurate detection even with large datasets.
Data Source
AI summary
A method, apparatus, and computer program product are provided for determining a location of on-street parking. The method includes receiving a plurality of individual probe points along a section of a roadway. Each individual probe point is associated with a location where each probe point was captured by one of one or more vehicles on the section of the roadway. The method also includes selecting one or more of the plurality of individual probe points based on a speed of the one or more vehicles associated with each probe point at a time the probe point was captured. The method further includes comparing the location of the selected probe points with a known roadway representation. Based on the comparison, the method also includes determining one or more groupings of a plurality of potential parking spaces along the section of the roadway. A corresponding apparatus and computer program product are provided.