Multilevel Feature Tracking With Direct Scale Selection for Motion Blur
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Motion blur in visual-inertial tracking systems degrades tracking performance and increases computational operations, particularly in AR/VR devices with fast camera movements, leading to inefficient power consumption and resource usage.
Innovation Solution
A method for predicting an optimal scale level for feature matching in image pyramids based on motion blur estimation and scale changes, reducing the need for multiple matching attempts and minimizing processing time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple scale levels are used for feature matching to handle motion blur, then tracking accuracy is improved, but computational complexity and processing time increase
Solution Approach 1:
The system performs preliminary estimation of motion blur parameters and scale changes before feature matching. Based on this preliminary analysis, the optimal scale level is predicted in advance, allowing the system to skip unnecessary matching attempts at other scales and directly proceed to the most promising scale level, thereby reducing overall computational complexity while maintaining accuracy.
Solution Approach 2:
The system incorporates feedback from motion blur estimation and scale change analysis to dynamically select the optimal scale level for feature matching. This feedback mechanism allows the system to adapt to varying motion conditions and avoid unnecessary computational operations at unsuitable scale levels, improving efficiency while maintaining tracking accuracy.
2Reliability
If multiple matching attempts are performed to handle varying motion blur conditions, then tracking robustness is improved, but processing time increases
Solution Approach 1:
The system performs preliminary estimation of motion blur and scale changes to predict the optimal scale level before initiating feature matching. This preliminary action eliminates the need for multiple sequential matching attempts, allowing the system to directly target the most promising scale level and thereby reduce processing time while maintaining robustness through informed scale selection.
Solution Approach 2:
The system changes the scale parameter dynamically based on motion blur estimation and scale change analysis. By adjusting the scale level prediction based on current motion conditions, the system can adapt to varying motion blur scenarios without requiring multiple fixed matching attempts, thus reducing processing time while maintaining tracking robustness across different motion conditions.
3Measurement precision
If computational operations are increased to maintain tracking accuracy under high dynamics, then tracking precision is improved, but power consumption increases
Solution Approach 1:
The system performs preliminary analysis of motion blur and scale changes to predict the optimal scale level before feature matching. This preliminary action enables the system to skip unnecessary computational operations at other scale levels, reducing overall processing requirements and power consumption while maintaining tracking precision through targeted computation at the predicted optimal scale.
Solution Approach 2:
Instead of performing complete feature matching at all scale levels, the system applies partial action by focusing computational resources only on the predicted optimal scale level. This selective approach reduces unnecessary computational operations and power consumption while maintaining sufficient tracking precision, avoiding the excessive action of processing all scale levels when not needed.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for mitigating motion blur in a visual-inertial tracking system is described. In one aspect, the method includes accessing a first image generated by an optical sensor of the visual tracking system, accessing a second image generated by the optical sensor of the visual tracking system, the second image following the first image, determining a first motion blur level of the first image, determining a second motion blur level of the second image, identifying a scale change between the first image and the second image, determining a first optimal scale level for the first image based on the first motion blur level and the scale change, and determining a second optimal scale level for the second image based on the second motion blur level and the scale change.