Parking Region Extraction Using Aerial Images and Map Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies struggle to create map information for parking lots when cameras and sensors are not installed, which is crucial for accurate navigation and parking guidance.
Innovation Solution
A parking region extraction apparatus and method that utilizes map information, aerial images, and notification data to identify and estimate parking regions by dividing sites, setting boundary lines, and calculating accuracy based on vehicle density and notification data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If cameras and sensors are installed in the parking lot to create map information, then the accuracy of parking region identification is improved, but the device complexity and cost increase
Solution Approach 1:
The patent uses aerial images as a copy or alternative representation of the parking lot surface, eliminating the need for physical cameras and sensors installed within the parking lot. The aerial images capture the entire parking region from above, providing the necessary visual information without requiring complex on-site detection devices.
Solution Approach 2:
The patent introduces map information as an intermediary data source that bridges the gap between aerial images and parking region identification. The map information provides contextual data about the parking lot layout, which is combined with aerial image analysis to accurately identify parking regions without direct sensor installation.
2Device complexity
If cameras and sensors are not installed in the parking lot, then the device complexity and cost are reduced, but the ability to create map information is lost
Solution Approach 1:
The patent replaces the need for on-site cameras and sensors by using aerial images as an alternative data source. These aerial images serve as a copy of the parking lot surface, providing sufficient information for map creation and parking region identification without requiring physical detection devices in the parking lot.
Solution Approach 2:
The patent substitutes the mechanical detection system (cameras and sensors installed in the parking lot) with a remote sensing approach using aerial images. This replacement eliminates the need for complex mechanical installation while maintaining the ability to create and update map information.
3Area of stationary object
If the site is divided into multiple candidates for parking regions, then the coverage is improved, but the difficulty of detecting and measuring the accurate parking region increases
Solution Approach 1:
The patent employs feedback mechanisms by comparing the detected parking regions with the provided map information. The map information serves as reference data that validates and refines the identified parking regions, allowing the system to verify accuracy and make corrections based on the comparison between detected features and expected features from the map.
Solution Approach 2:
The patent divides the parking lot into multiple candidate regions based on aerial image analysis, then uses map information to evaluate and select the most accurate parking regions. This segmentation approach allows comprehensive coverage while using external reference data to simplify the accuracy verification process for each segment.
Data Source
AI summary
A parking region extraction apparatus is a parking region extraction apparatus for extracting a parking region which is a region for parking a vehicle in a site, and includes an acquisition unit that acquires map information including the site, an extraction unit that extracts candidates for the parking region by dividing the site using the map information, and an estimation unit that derives an accuracy of each of the candidates and estimate the parking region.