Probe Data Map-Matching Classification via Density Histograms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods fail to accurately distinguish between map-matched and non-map-matched probe data points, which can introduce bias in analysis and hinder the effective use of probe data for route planning and navigation services.
Innovation Solution
A method involving a processing system that generates a probe density histogram to classify probe data points as map-matched or non-map-matched based on their distribution across a road segment, using statistically significant deviations and confidence values to determine the accuracy of map-matching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If probe data points are processed without distinguishing map-matched vs non-map-matched status, then processing simplicity is maintained, but analysis accuracy deteriorates due to introduced bias
Solution Approach 1:
The patent segments probe data points into two distinct categories: map-matched data points and non-map-matched data points. This segmentation is achieved by analyzing the distribution of data points across road segments and identifying statistical deviations that indicate map-matching has been applied. By separating these categories, the system enables different processing approaches for each type, thereby improving analysis accuracy while maintaining reasonable processing complexity.
2Measurement precision
If map-matching is applied to all probe data points, then location accuracy is improved, but data authenticity deteriorates due to loss of original position information
Solution Approach 1:
The patent introduces an intermediary analysis step that examines the distribution characteristics of probe data points without immediately applying map-matching transformation. By using statistical analysis of data point distributions across road segments as an intermediary, the system can identify which points have been map-matched and which retain original positioning, thereby preserving information about the original state while still enabling accurate location analysis.
3Reliability
If probe data points are classified by map-matching status, then analysis reliability is improved, but processing complexity increases due to additional classification steps
Solution Approach 1:
The patent implements a self-service classification mechanism where the map-matching status of data points is determined through self-analysis of their distribution characteristics. The system automatically identifies map-matched versus non-map-matched points by examining statistical deviations in their spatial distribution across road segments, eliminating the need for external manual classification or complex preprocessing steps while maintaining high analysis reliability.
Data Source
AI summary
A method, apparatus, and computer program product are provided for determining whether probe data points are map-matched or non-map-matched such that the probe data can be processed and analyzed appropriately without introducing bias into the analysis which may be caused by map-matching ahead of analysis. A mapping system includes a memory having map data stored therein, and processing circuitry. The processing circuitry may be configured to receive probe data points associated with a plurality of vehicles. Each probe data point is received from a probe apparatus of a plurality of probe apparatuses. The probe apparatus includes a plurality of sensors and being onboard a respective vehicle. Each probe data point includes location information associated with the respective probe apparatus.