Image Georegistration Using 3D Model Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing augmented reality systems face inaccuracies in georegistration due to limitations in measurement systems, leading to incorrect superimposition of AR content in relation to environmental features, especially under varying environmental conditions and over time.
Innovation Solution
A method and system that utilize a 3D geographic model to correct inaccuracies in sensor location and orientation by comparing sensor images with virtual images from the model, allowing for accurate georegistration and correct placement of AR content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If measurement systems (GPS, magnetometer, compass, accelerometer) are used to determine camera position and orientation, then the system can provide location and orientation data, but the accuracy deteriorates under varying environmental conditions and over time
Solution Approach 1:
The system captures sensor images, generates corresponding model images from a 3D geographic model using the measured position and orientation, compares the two images to determine discrepancies, and uses this feedback to calculate corrected position and orientation coordinates. This closed-loop feedback mechanism continuously refines the accuracy of location data regardless of environmental conditions.
Solution Approach 2:
The patent creates a virtual copy of the real-world scene by generating a model image from a 3D geographic model using the measured camera parameters. This virtual model image serves as a reference to compare against the actual sensor image, enabling detection and correction of measurement inaccuracies without being affected by environmental degradation.
2Productivity
If AR content is superimposed based on measured camera position and orientation, then the system can display augmented reality content, but the content is misplaced relative to environmental features due to measurement inaccuracies
Solution Approach 1:
The system uses image comparison between sensor captures and model-generated images to generate feedback about position and orientation discrepancies. This feedback is used to correct the coordinates before AR content is superimposed, ensuring accurate placement of virtual elements relative to real-world environmental features even when initial measurements are inaccurate.
Solution Approach 2:
The patent performs preliminary correction of camera position and orientation coordinates by comparing sensor images with model images before the AR content superimposition step. This preliminary action ensures that the coordinate system is already corrected when AR content is placed, preventing misalignment issues.
3Measurement precision
If a 3D geographic model is used to correct position and orientation, then measurement accuracy is improved, but the system complexity increases
Solution Approach 1:
The 3D geographic model serves multiple functions: it provides the reference framework for generating model images, acts as the basis for comparing with sensor images to detect discrepancies, and enables calculation of corrected position and orientation coordinates. This multi-functionality reduces the need for separate correction devices or systems.
Solution Approach 2:
The system uses a virtual copy (model image) of the real scene from the 3D geographic model to compare against the actual sensor image. This copying approach allows accuracy improvement through computational comparison rather than adding complex physical measurement devices, reducing overall system complexity while maintaining high precision.
Data Source
Figure 1
Figure 2
Figure 3A~3B
AI summary
Image georegistration method and system. An imaging sensor acquires a sensor-image of a scene. Imaging parameters of the acquired sensor-image are obtained, the imaging parameters including at least the detected 3D position and orientation of the imaging sensor when acquiring the sensor-image, as detected using a location measurement unit. A model-image of the scene is generated from a textured 3D geographic model, the model-image representing a texture-based 2D image of the scene as acquired in the 3D model using the imaging parameters. The sensor-image and the model-image are compared and the discrepancies between them determined. An updated 3D position and orientation of the imaging sensor is determined in accordance with the discrepancies. The updated position and orientation may be used to display supplementary content overlaid on the sensor-image in relation to a selected location on the sensor-image, or for determining the geographic location coordinates of a scene element.