Parking Map Refinement via Anomaly Distribution Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current parking area mapping technologies fail to dynamically identify and update anomalies in parking regions, such as obstacles or restricted areas, which can lead to inaccuracies in available parking space detection.
Innovation Solution
A method involving processing circuitry that receives data from vehicles driving through parking regions, calculates distributions of detected objects, identifies differences between parking region and street segment distributions using statistical tests, and generates an updated parking area map to highlight anomalies, allowing for real-time refinement of parking maps.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If parking area maps are created using sensor data from vehicles, then parking spaces can be detected, but anomalies such as obstacles or restricted areas cannot be identified, leading to inaccurate parking space detection
Solution Approach 1:
The patent introduces an intermediary anomaly detector that compares object distributions between parking regions and street segments to identify anomalies. This mediator system processes sensor data to distinguish between actual parking spaces and obstacles, resolving the contradiction between detecting parking spaces and identifying anomalies.
Solution Approach 2:
The patent segments the analysis into two distinct distributions: parking region distribution and street segment distribution. By comparing these segmented distributions, the system can identify anomalies that differ from normal parking patterns, thereby improving both parking space detection accuracy and anomaly identification capability.
2Productivity
If static parking maps are used, then map creation is simple, but the maps cannot be dynamically updated to reflect current parking availability and anomalies
Solution Approach 1:
The patent implements a feedback mechanism where sensor data from vehicles continuously flows back to update the parking map. The anomaly detector processes this feedback data to identify changes in parking regions, enabling dynamic updates that maintain current parking status information while improving map update speed.
Solution Approach 2:
The patent transforms the static parking map into a dynamic system that adapts to changing conditions. By continuously comparing object distributions and updating the map based on detected anomalies, the system maintains real-time accuracy without requiring complete remapping, thus improving productivity while preserving information quality.
3Adaptability or versatility
If all detected objects are classified as parking spaces, then parking space detection coverage is maximized, but obstacles and restricted areas are misclassified, reducing detection accuracy
Solution Approach 1:
The patent applies local quality by treating different regions with different classification rules. Parking regions are analyzed for potential parking spaces, while street segments are analyzed for anomalies. This localized approach allows maximum detection coverage in parking areas while maintaining high classification accuracy by identifying and excluding anomalies from parking space classification.
Data Source
AI summary
A method includes receiving data corresponding to detected objects in a parking region, the data including information ascertained by an ascertaining vehicle driving through the parking region, determining a parking region distribution, based on detected objects in a street segment located within the parking region, determining a street segment distribution, calculating a difference between the parking region and the street segment distribution, identifying an anomaly located within the street segment, and generating an updated parking area map of the parking region based on the identified anomaly. The method includes receiving the data, for example, each time an ascertaining vehicle drives through the parking region.