Event Camera Gaze Tracking Using Coded Glint Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current three-dimensional (3D) gaze tracking methods require additional hardware and software to extract pupil information, leading to inefficiencies and high costs due to the complexity of processing eye movements for human-computer interaction.
Innovation Solution
A method using an event camera to capture glint information from light sources with unique emission patterns, estimating the corneal sphere center and eye rotation center positions to determine 3D gaze-related information without relying on pupil detection, thereby reducing computational load and costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional 3D gaze tracking methods extract and use pupil information, then gaze tracking accuracy is improved, but additional hardware and software are required, increasing device complexity and cost
Solution Approach 1:
The patent extracts and utilizes only the necessary glint information from light reflections on the eye surface, eliminating the need for complex pupil detection systems. By focusing solely on glint position data from multiple light sources, the system achieves adequate gaze tracking without the additional hardware and software required for full pupil information extraction.
Solution Approach 2:
The patent employs multiple inexpensive light sources with unique codes instead of complex imaging systems. These light sources emit coded patterns that can be detected by a simple camera, replacing expensive pupil-tracking hardware with a more economical solution using basic optical components and computational processing.
2Loss of information
If traditional 3D gaze tracking methods process pupil information, then comprehensive eye movement data is obtained, but computational load increases, reducing processing efficiency
Solution Approach 1:
The patent extracts only the essential glint position information from light reflections, ignoring other eye parameters such as pupil size and shape. This selective extraction of minimal necessary data significantly reduces computational load while maintaining sufficient information for accurate gaze direction estimation.
Solution Approach 2:
The patent uses multiple light sources with unique codes, emitting more light signals than strictly necessary, to simplify the detection process. This excessive action in terms of light emission enables the system to process gaze information more efficiently by providing redundant, easily distinguishable signals that reduce processing complexity.
3Reliability
If additional hardware and software are used for pupil information extraction, then gaze tracking capability is enhanced, but system cost increases
Solution Approach 1:
The patent makes the camera serve multiple functions: it detects both the glint positions from light sources and provides sufficient data for gaze tracking. This multi-functionality eliminates the need for separate specialized hardware, reducing system cost while maintaining reliable gaze tracking capability through clever use of existing components.
Solution Approach 2:
The patent introduces coded light sources as intermediaries between the eye and the camera. These light sources emit unique patterns that mediate the interaction, allowing the simple camera to extract gaze information without requiring complex imaging systems. The coded light acts as a mediator that simplifies the overall system architecture.
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 enables efficient and cost-effective 3D gaze tracking by eliminating the need for additional hardware and software, achieving low-latency performance suitable for applications like augmented reality systems.
Implementation Method 1
obtaining glint information based on capturing, using an event camera, a glint signal corresponding to light emitted from a plurality of light sources
Data Source
AI summary
An electronic device may operate a plurality of light sources, where each light source operates according to a light source code of a light source code set, each light source code being unique with respect to each other light source code, capture a glint signal corresponding to light emitted from the plurality of light sources through an event camera, obtain glint information from event data from the event camera, estimate a corneal sphere center position and an eye rotation center position based on the glint information, and determine three-dimensional (3D) gaze-related information based on the corneal sphere center position and the eye rotation center position.