Map Data Generation Using Dual-Resolution Image Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies for autonomous driving face challenges in accurately generating and utilizing map data to determine the surrounding environment of vehicles, requiring high-resolution image data and precise location information to ensure reliable navigation and control.
Innovation Solution
An electronic device that obtains image data of different resolutions while moving, uses high-resolution data to obtain accurate pose information and location data, and generates map data by matching this information with low-resolution image data, allowing for accurate location estimation without the need for expensive high-resolution sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-resolution image data is used to generate map data for autonomous driving, then location estimation accuracy is improved, but device cost increases
Solution Approach 1:
The patent divides the image data into two resolution levels: high-resolution data for accurate location estimation and low-resolution data for map generation. This segmentation allows the system to use only high-resolution data when needed for positioning, while relying on lower-resolution cached map data for other purposes, thereby reducing the need for continuously capturing and processing expensive high-resolution images.
Solution Approach 2:
The patent applies different quality levels of image data to different functional requirements. High-resolution data is used locally and specifically for location estimation where precision is critical, while low-resolution data is used for general map representation and navigation where such precision is not required. This local quality differentiation reduces overall system cost while maintaining necessary accuracy.
2Reliability
If high-resolution sensors are deployed for autonomous driving, then map data accuracy is improved, but device complexity increases
Solution Approach 1:
The patent implements a dynamic system that adapts the use of high-resolution sensors based on operational needs. The system activates high-resolution imaging only when location estimation accuracy is critical, while relying on pre-generated low-resolution map data during normal operation. This dynamic approach maintains map data accuracy when needed while reducing sensor complexity and computational burden during routine navigation.
Solution Approach 2:
The patent performs preliminary generation of low-resolution map data in advance, which can be stored and reused. This preliminary action reduces the need for real-time high-resolution data processing during autonomous driving operations, thereby reducing device complexity while maintaining the ability to generate accurate maps when required.
3Ease of manufacture
If low-resolution image data is used for map generation, then device cost is reduced, but location estimation accuracy deteriorates
Solution Approach 1:
The patent introduces a matching mechanism as an intermediary process that compares low-resolution current images with high-resolution cached map data. This intermediary matching process enables the system to achieve accurate location estimation by finding correspondences between features in low-resolution images and their high-resolution representations in the cached map, thereby overcoming the limitation of using low-resolution sensors.
Data Source
AI summary
Provided are an electronic device for generating map data and an operating method of the electronic device. The operating method of the electronic device includes: obtaining image data of a first resolution and image data of a second resolution for each of a plurality of nodes generated while the electronic device moves; obtaining location information with respect to each of the generated nodes, by using the image data of the second resolution; generating and storing map data by matching the obtained location information with the image data of the first resolution for each node; and estimating a current location of the electronic device by using the generated map data and the image data of the first resolution.