Virtual Selfie Stick Imaging With Selective Blur-Aware Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing augmented and virtual reality devices experience high motion blur during fast movements, leading to degraded tracking performance and increased computational demands.
Innovation Solution
A method for generating a virtual selfie stick image by partitioning the image into invisible, blocked, and interpolatable areas, using computer vision algorithms like Multiple View Stereo (MVS) or Neural Radiance Fields (NeRF) for the former two areas and interpolation for the latter, reducing computational load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If motion tracking uses images captured by optical sensor during fast movement, then tracking function is enabled, but motion blur degrades tracking performance
Solution Approach 1:
The patent segments the image processing task by dividing the captured blurry image into multiple regions (e.g., foreground objects, background, motion-affected areas). Different processing strategies are applied to different regions: some regions undergo deblurring algorithms while others are processed differently, allowing selective improvement of tracking-relevant areas without processing the entire image uniformly.
Solution Approach 2:
The patent applies preliminary deblurring or enhancement operations to the captured image before feeding it to the motion tracking algorithm. By pre-processing the image to reduce motion blur effects, the tracking system receives higher quality input data, improving tracking reliability without requiring changes to the core tracking methodology.
2Measurement precision
If computational operations are increased to maintain tracking accuracy under high dynamics, then tracking accuracy is maintained, but processor cycles and power consumption increase
Solution Approach 1:
The patent applies computational resources selectively to specific regions of the image rather than processing the entire image uniformly. High-computation deblurring algorithms are applied only to regions containing tracking-relevant features, while other regions receive minimal or no processing. This localizes computational effort to where it is most needed, reducing overall power consumption while maintaining tracking accuracy.
Solution Approach 2:
The patent implements a partial processing approach where only the necessary portions of the image are processed at high computational levels. Instead of applying full deblurring algorithms to the entire image, the system processes only the minimum required regions to achieve adequate tracking accuracy, thereby reducing unnecessary computational overhead and power consumption.
Data Source
AI summary
A method for generating a virtual selfie stick image is described. In one aspect, the method includes generating, at a device, an original self-portrait image with an optical sensor of the device, the optical sensor directed at a face of a user of the device, the device being held at an arm length from the face of the user, displaying, on a display of the device, an instruction guiding the user to move the device at the arm length about the face of the user within a limited range at a plurality of poses, accessing, at the device, image data generated by the optical sensor at the plurality of poses, and generating a virtual selfie stick self-portrait image based on the original self-portrait image and the image data.