3D Image Tile Updating Using Vehicle-Captured Deviation Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for generating and updating 3D map tiles of transportation networks face challenges in maintaining accuracy and timeliness, especially when active cameras are not available, as they rely on periodic updates that may not reflect real-time changes or recent visual information.
Innovation Solution
A method and system that utilize image deviation data to determine when image tiles need updating, scheduling vehicles with sensors to capture new data or identifying passing vehicles to refresh tiles, ensuring that the map remains current by prioritizing updates based on image tile changes and using metadata and machine learning for efficient data collection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If periodic updates are used to maintain map tiles, then system complexity is reduced, but map accuracy and timeliness deteriorate
Solution Approach 1:
The system implements feedback by continuously monitoring image deviation data from sensors on moving vehicles and using this information to trigger selective updates of map tiles only when changes are detected, rather than using fixed periodic updates. This maintains map accuracy while avoiding unnecessary system complexity.
Solution Approach 2:
The map update system serves itself by automatically detecting when updates are needed through image deviation analysis and autonomously scheduling vehicle routes to capture fresh imagery only for affected areas, eliminating the need for external manual intervention or complex centralized scheduling.
2Loss of time
If active cameras are deployed in all areas, then map update timeliness is improved, but system cost and complexity increase
Solution Approach 1:
The system makes moving vehicles serve multiple functions: they perform their primary transportation role while simultaneously acting as mobile sensing platforms that capture imagery for map updates. This eliminates the need for dedicated stationary cameras in all areas, reducing system complexity while maintaining update timeliness.
Solution Approach 2:
The system introduces image deviation data as an intermediary that mediates between the need for timely updates and the availability of sensing resources. By analyzing this intermediate data, the system intelligently determines which areas need updates and schedules vehicle passes accordingly, avoiding the need for universal camera deployment.
3Loss of information
If full image tiles are captured every time, then data completeness is improved, but data transmission and processing load increase
Solution Approach 1:
The system extracts only the necessary information by capturing and transmitting solely the portions of image tiles that have changed, as identified through image deviation data analysis. This eliminates unnecessary data transmission and processing while maintaining complete information about areas that actually require updates.
Solution Approach 2:
The system applies partial action by capturing only the minimal necessary data - specifically, only those image tile portions that show deviation from the current map. This avoids the excessive action of capturing complete full-resolution tiles every time, reducing data volume while maintaining sufficient completeness for update purposes.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method may include examining image deviation data of image tiles of a set that form a set view of a volume of space and determining whether the tiles are due for revision to update the tiles. The method may include scheduling a first vehicle (102) to capture updated image tiles and/or identifying a second vehicle (102) to capture the updated image tiles. A system (150) may include one or more processors (132) to examine image deviation data of image tiles of a set that form a set view of a volume of space and determine whether the one or more image tiles of the set are due for revision to update the view. The one or more processors (132) schedule a first vehicle (102) to move through or by the volume of space to capture one or more updated image tiles or identify a second vehicle (102) to capture the one or more updated image tiles.