Eye Tracking Illuminator Shutdown for Reflection Blob Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Eye tracking systems in VR devices face inaccuracies due to reflections from spectacles or other optical assemblies, making it difficult to determine eye direction and gaze direction reliably.
Innovation Solution
An eye tracking system with multiple illuminators and a processing circuitry that identifies and switches off the active illuminator causing reflections by applying different criteria for blob detection before and after illuminator shutdown, using region-based analysis to minimize processing load and power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple illuminators are used to illuminate the eye for eye tracking, then the quality of eye images and tracking accuracy is improved, but unwanted reflections from optical arrangements (such as spectacles) are generated that reduce measurement precision
Solution Approach 1:
The patent extracts and removes the harmful reflection (blob) from the eye image by identifying it as a separate detectable object and excluding it from the eye tracking calculation. The processing circuitry detects blobs that do not move with eye movement and removes their influence from the gaze direction determination.
Solution Approach 2:
The system uses feedback by detecting the presence and position of blobs in real-time during image acquisition, and dynamically adjusting the eye tracking calculation based on this detected information. The blob detection results feed back into the gaze estimation process to compensate for the harmful reflections.
2Reliability
If traditional blob detection methods are used to identify reflections, then the system complexity is reduced, but the reliability of eye tracking is insufficient when blobs are present
Solution Approach 1:
The patent segments the image processing task into distinct stages: initial blob detection using first criteria, verification using second criteria, and conditional processing based on blob presence. This segmentation allows the system to apply different processing levels based on the specific situation, improving reliability without always requiring full complexity.
Solution Approach 2:
The system dynamically adjusts its processing behavior based on blob detection results. When blobs are detected, the system activates additional verification steps and modifies the eye tracking calculation. When no blobs are present, the system uses the standard simpler process, thus adapting complexity to actual needs.
3Illumination intensity
If all illuminators are kept on to ensure sufficient eye illumination, then the image quality is maintained, but power consumption increases and unwanted reflections persist
Solution Approach 1:
The patent implements periodic action by switching illuminators on and off based on detection cycles. The processing circuitry detects blobs in periodic intervals and controls illuminator activation accordingly, turning off illuminators when not needed and turning them back on when eye tracking requires illumination, thus reducing overall power consumption while maintaining functionality.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enhances the accuracy of eye tracking by reducing false correlations between illuminators and reflections, improving tracking reliability and efficiency.
Implementation Method 1
the first image resulting from the image sensor detecting light from the plurality of illuminators reflected from the eye of the user and reflected from an optic arrangement located between the plurality of illuminators and the eye of the user
Data Source
AI summary
An eye tracking system is provided that detects the presence of problematic blobs in an image captured by the system and removes these problematic blobs by switching off illuminators. Problematic blobs may be those obscuring the pupil of the eye of the user. Each blob is detected in a first image by the use of at least one first criterion, and then an illuminator is switched off. After the illuminator is switched off, at least one second criterion is used to identify blobs in a subsequent image. This process may be repeated until the illuminator causing the problematic blob is identified.