Crowdsourced 3D Model Updates via Visual Difference Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for creating and maintaining real-world 3D models are costly and time-consuming, and these models become outdated due to changes at the location over time, as they rely on large-scale professional data collection efforts.
Innovation Solution
A scalable method using crowd-sourced imaging data, where users capture video segments with positioning and orientation information, which are compared to existing 3D geometry projections to detect visual differences, allowing for updating the model only when changes are below a certain threshold, thereby maintaining an accurate and up-to-date representation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If large-scale professional data collection efforts are used to create and maintain 3D models, then the accuracy and completeness of the models are improved, but the cost and time required increase significantly
Solution Approach 1:
The system enables automatic self-updating of 3D models by comparing newly captured images with existing model projections. The computer automatically detects changes, determines whether updates are needed based on threshold comparisons, and integrates new data without requiring manual intervention, thus reducing time and cost while maintaining accuracy
Solution Approach 2:
The system creates projections of the existing 3D geometry and compares them with newly captured images. By working with projected copies rather than directly manipulating the full 3D model, the system efficiently detects changes and updates only necessary portions, reducing processing time and computational resources
2Measurement precision
If large-scale professional data collection efforts are used to create and maintain 3D models, then the accuracy and completeness of the models are improved, but the cost increases significantly
Solution Approach 1:
The system performs automatic change detection and model updating through algorithmic comparison of images and projections. This automation eliminates the need for expensive professional data collection teams, reducing costs while maintaining model accuracy through systematic automated processes
Solution Approach 2:
The system updates only the portions of the 3D model that have actually changed, rather than performing complete re-collection and re-processing. By detecting change regions and updating only those areas, the system reduces computational cost and resource requirements while maintaining overall model accuracy
3Reliability
If 3D models are updated frequently to reflect location changes, then the model currency is improved, but the processing complexity increases
Solution Approach 1:
The system extracts and compares only the specific regions where changes have occurred between the new images and model projections. By isolating change regions rather than processing entire models, the system reduces processing complexity while ensuring timely updates to maintain model currency
Solution Approach 2:
The system applies different processing strategies to different parts of the model based on local change characteristics. Change regions undergo detailed analysis and selective updating, while unchanged regions are left untouched, optimizing processing efficiency and reducing overall complexity
4Productivity
If threshold-based filtering is applied to determine when to update the model, then unnecessary updates are reduced, but some relevant changes might be missed
Solution Approach 1:
The system uses threshold-based filtering as a feedback mechanism to determine whether detected changes warrant model updates. By comparing change magnitude against thresholds, the system efficiently filters out noise while capturing significant changes, balancing update efficiency with detection accuracy
Data Source
AI summary
An exemplary method includes prompting a user to capture video data at a location. The location is associated with navigation directions for the user. Information representing visual orientation and positioning information associated with the captured video data is received by one or more computing devices, and a stored data model representing a 3D geometry depicting objects associated with the location is accessed. Between corresponding images from the captured video data and projections of the 3D geometry, one or more candidate change regions are detected. Each candidate change region indicates an area of visual difference between the captured video data and projections. When it is detected that a count of the one or more candidate change regions is below a threshold, the stored model data is updated with at least part of the captured video data based on the visual orientation and positioning information associated with the captured video data.