Hybrid AR Tracking Using Watermark Pose and Feature Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing augmented reality systems face challenges in determining the pose of a camera relative to a tracking image without pre-loaded feature point data and require pre-processing of tracking images, leading to limitations in real-time augmentation and communication dependencies.
Innovation Solution
The system uses a combination of digital watermarks and feature points, where the initial pose is determined using watermark information, and subsequent tracking relies on feature points, allowing for on-the-fly extraction and updating of feature points, with periodic or opportunistic watermark-based recalibration for accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If feature point data is pre-loaded for tracking, then tracking accuracy is improved, but device memory requirements and initial processing time increase
Solution Approach 1:
The system performs preliminary extraction of feature points from the tracking image and stores them in a data structure before tracking begins. This pre-extraction eliminates the need for real-time feature extraction during tracking, improving both accuracy and efficiency without requiring pre-loaded data from external sources
Solution Approach 2:
The system extracts feature points directly from the captured tracking image using onboard processing capabilities, rather than relying on pre-loaded feature point data from external sources. This self-service approach reduces memory requirements while maintaining tracking accuracy
2Quantity of substance
If feature points are extracted in real-time during tracking, then memory usage is reduced, but processor cycles and battery drain increase
Solution Approach 1:
Feature points are extracted from the tracking image in advance and stored in a data structure before the tracking process begins. This preliminary extraction eliminates the need for continuous real-time feature extraction during tracking, significantly reducing processor cycles and battery consumption while maintaining real-time tracking capability
Solution Approach 2:
The system performs feature point extraction periodically or at specific intervals rather than continuously during tracking. This periodic approach reduces the overall computational load and energy consumption while still providing sufficient feature data for accurate real-time tracking
3Productivity
If watermark-based pose determination is used initially, then real-time tracking can begin without pre-loaded data, but periodic recalibration increases processing overhead
Solution Approach 1:
The system uses watermark information to determine initial camera pose relative to the tracking image before feature point tracking begins. This preliminary pose determination enables real-time tracking to start immediately without requiring pre-loaded feature point data from external sources
Solution Approach 2:
The system implements periodic recalibration using watermark-based pose determination to verify and correct feature point tracking accuracy. This feedback mechanism ensures continued accuracy while minimizing processing overhead by recalibrating only when necessary rather than continuously
4Measurement precision
If pre-processing of tracking images is performed, then tracking accuracy is improved, but communication dependencies and setup time increase
Solution Approach 1:
The system performs all necessary image processing and feature point extraction operations locally on the captured tracking image using onboard processing capabilities. This eliminates communication dependencies on external servers for data provision while maintaining tracking accuracy through self-contained processing
Solution Approach 2:
The system extracts feature points and prepares tracking data structures from the captured image in real-time before augmentation begins, eliminating the need for pre-processing and data provision in advance. This on-the-fly preparation reduces setup time while maintaining accuracy
Data Source
AI summary
A camera captures video imagery depicting a digitally-watermarked object. A reference signal in the watermark is used to discern the pose of the object relative to the camera, and this pose is used in affine-transforming and positioning a graphic on the imagery as an augmented reality overlay. Feature points are also discerned from the captured imagery, or recalled from a database indexed by the watermark. As the camera moves relative to the object, the augmented reality overlay tracks the changing object depiction, using these feature points. When feature point-based tracking fails, the watermark is again processed to determine pose, and the overlay presentation is updated accordingly. In another arrangement, feature points are extracted from images of supermarket objects captured by multiple users, and are compiled in a database in association with watermark data identifying the objects—serving as a crowd-sourced repository of feature point data. A great number of other features and arrangements are also detailed.