Index Reliability Weighting for Position Orientation Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional mixed reality registration techniques are prone to errors due to detection errors and low precision in index detection, leading to unstable position and orientation measurements, especially when there are a small number of detected indices or under varying illumination conditions.
Innovation Solution
A method and apparatus for position and orientation measurement that input sensed images, detect indices, calculate reliability based on detection conditions, and weight indices to reduce the influence of erroneous or low-precision detections, using a reliability calculation unit to improve registration precision and stability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional registration techniques use all detected indices with uniform weights, then the calculation process is simple, but measurement precision deteriorates due to large influence of erroneous or low-precision indices
Solution Approach 1:
The patent applies local quality by assigning different weights to different indices based on their individual detection conditions and reliabilities. Instead of uniform weighting, each index is evaluated locally according to its specific detection quality, allowing high-precision indices to have greater influence while erroneous or low-precision indices are downweighted or eliminated from the calculation.
Solution Approach 2:
The patent changes the parameter of index reliability by introducing detection condition analysis and reliability calculation. The system dynamically adjusts the weight parameter for each index based on detected conditions such as detection precision, illumination quality, and detection history, thereby transforming the static uniform weighting approach into a dynamic adaptive weighting system.
2Reliability
If statistical estimation methods are used to eliminate erroneous indices, then measurement reliability improves, but the method is not satisfactory when there are a small number of indices or under varying illumination conditions
Solution Approach 1:
The patent applies preliminary action by pre-establishing a comprehensive evaluation system for index detection conditions before performing the actual position and orientation calculation. The system pre-calculates reliability metrics for each index based on detection conditions, illumination quality, and historical detection data, so that when indices are scarce or conditions vary, the pre-computed reliability information is already available to guide the selection and weighting of usable indices.
Solution Approach 2:
The patent introduces dynamics by making the index selection and weighting process adaptive rather than static. The system continuously evaluates detection conditions and adjusts the reliability and weight of each index dynamically based on current illumination conditions, detection quality, and historical performance. This dynamic adaptation allows the system to maintain reliability whether there are many or few indices, and whether illumination conditions are stable or varying.
3Measurement precision
If uniform weights are applied to all indices, then calculation complexity is low, but position and orientation measurement becomes unstable due to detection errors
Solution Approach 1:
The patent applies local quality by assigning different weights to different indices based on their individual detection conditions and reliabilities. Instead of uniform weighting, each index is evaluated locally according to its specific detection quality, allowing high-precision indices to have greater influence while erroneous or low-precision indices are downweighted or eliminated from the calculation.
Solution Approach 2:
The patent changes the parameter of index reliability by introducing detection condition analysis and reliability calculation. The system dynamically adjusts the weight parameter for each index based on detected conditions such as detection precision, illumination quality, and detection history, thereby transforming the static uniform weighting approach into a dynamic adaptive weighting system.
Data Source
AI summary
Indices allocated or set on an object are detected from sensed images of frames, which are time-serially sensed by an image sensing device. In association with each index detected from the sensed image of the frame of interest, a detection condition from frames before the frame of interest is acquired. According to the acquired detection condition of the index, the reliability of this index is calculated. The position and orientation of the object or the image sensing device are calculated using at least information associated with the image coordinates of the detected indices and the reliabilities of the indices.