3D SPAD Eye Tracking With Compressive Sensing Depth Capture
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing eye tracking technologies in near-eye display devices face challenges with high speed measurement, limited bandwidth, low resolution, and noise susceptibility, particularly with single photon avalanche diode (SPAD) sensors, which are difficult to implement in augmented and virtual reality applications due to strict size and power constraints.
Innovation Solution
A three-dimensional (3D) compressive sensing based eye tracking system using SPAD sensors projects active encoded illumination patterns onto the eye, captured by synchronized high-speed illuminators like DMDs or MEMS, and employs compressive sensing techniques to achieve high resolution depth measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If traditional high-speed cameras are used for eye tracking, then measurement speed is improved, but device complexity and power consumption increase
Solution Approach 1:
The patent replaces traditional mechanical high-speed camera systems with a photonic solution using SPAD sensors and optical encoding. The system uses projected light patterns and photon detection instead of mechanical imaging components, achieving high-speed measurement while reducing device complexity and power consumption.
Solution Approach 2:
The patent changes the measurement parameters by using time-correlated single photon counting and encoding light patterns instead of traditional frame-based imaging. This parameter transformation enables ultra-fast measurement rates (microsecond scale) without requiring complex high-speed camera hardware.
2Use of energy by moving object
If SPAD sensors are used for eye tracking, then power consumption is reduced, but measurement precision deteriorates due to noise susceptibility
Solution Approach 1:
The patent introduces an optical encoding intermediary (projected light patterns) between the illumination source and the SPAD sensor. This intermediary carries encoded information that enables precise depth and gaze measurement while allowing the low-power SPAD sensor to operate in photon-counting mode, effectively filtering out ambient noise and improving measurement precision.
Solution Approach 2:
The patent uses periodic modulation of the projected light patterns at specific frequencies. This periodic action enables the SPAD sensor to distinguish between the encoded measurement light and ambient noise through frequency discrimination, thereby improving measurement precision while maintaining low power consumption.
3Ease of manufacture
If conventional eye tracking methods are used, then implementation is simpler, but measurement precision deteriorates due to limited bandwidth
Solution Approach 1:
The patent transitions from 2D image-based eye tracking to 3D depth measurement by introducing the time dimension through time-correlated single photon counting. This dimensional transformation enables precise depth measurement of the eye surface while maintaining implementation simplicity through the use of standard SPAD sensor technology.
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
This approach enhances eye tracking accuracy and power efficiency while reducing computational resources, enabling ultra-fast gaze tracking without the need for additional high-speed cameras.
Implementation Method 1
single photon avalanche diode (SPAD) sensors
Implementation Method 2
time-resolved information
Implementation Method 3
projection of active encoded illumination patterns onto a surface of an eye and capturing a reflection of the projected pattern
Data Source
AI summary
A three-dimensional (3D) compressive sensing based eye tracking system using single photon avalanche diode (SPAD) sensors achieves high resolution depth measurement by using low resolution SPAD sensors and active encoded illumination such as two- or three-dimensional fringe patterns, random speckles, random patterns, and/or superimposed patterns projected onto a surface of an eye. The patterns may be projected by high speed illuminators such as a digital micromirror device (DMD) or micro-electromechanical system (MEMS) projector in series changing at the same rate as the SPAD sensor capture rate. A processor may employ compressive sensing techniques to obtain a high resolution image and depth information from the captured images of varying patterns.


