Eye Tracking Using Structured Light for VR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing eye tracking techniques for virtual reality and augmented reality applications, such as video-based pupil tracking, require substantial computing resources, are susceptible to occlusion, and depend on iris-pupil contrast, making them inefficient and inaccurate, especially in head-mounted displays where space and resource constraints are significant.
Innovation Solution
An eye tracking unit utilizing structured light emitters and cameras to estimate the orientation of the eye by detecting distortions in a projected light pattern, allowing for accurate determination of the eye's orientation, including yaw, pitch, roll, and translation, without relying on iris-pupil contrast, and can be integrated into head-mounted displays for efficient gaze direction tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If video-based pupil tracking is used, then eye orientation can be detected, but substantial computing resources are required
Solution Approach 1:
The patent replaces the complex video-based machine vision algorithm with a structured light projection system. By projecting known light patterns onto the eye and capturing the reflected patterns with a camera, the system converts the computational problem of pupil detection into a geometric problem of pattern matching, significantly reducing computing resource requirements while maintaining measurement precision.
Solution Approach 2:
The patent uses structured light patterns as optical copies or projections of known geometric structures onto the eye surface. By comparing the projected pattern with the captured reflected pattern, the system determines eye orientation through geometric relationships rather than complex image analysis, reducing computational complexity.
2Measurement precision
If video-based pupil tracking is used, then eye orientation can be detected, but the method is susceptible to occlusion by eyelashes and eyelids
Solution Approach 1:
The patent illuminates specific local regions of the eye with structured light patterns, primarily targeting the cornea and sclera surfaces. By focusing on these local areas that are less susceptible to occlusion compared to the pupil region, the system maintains reliable eye orientation detection even when eyelashes or eyelids partially block the view of the pupil.
Solution Approach 2:
The patent divides the eye surface into multiple regions and projects structured light patterns onto specific segments (cornea, sclera) that provide sufficient geometric information for orientation detection. This segmentation approach allows the system to bypass occluded regions and use alternative surface features for accurate tracking.
3Measurement precision
If video-based pupil tracking is used, then eye orientation can be detected, but the method depends on iris-pupil contrast which varies across users
Solution Approach 1:
The patent changes the fundamental measurement parameter from iris-pupil contrast (which varies across users) to the geometric distortion of structured light patterns reflected from the eye surface. This parameter change makes the measurement independent of user-specific eye characteristics, improving adaptability and versatility across different users while maintaining measurement precision.
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
The solution provides accurate and efficient eye tracking in virtual reality and augmented reality environments, reducing computational requirements and overcoming occlusion issues, enabling precise gaze direction determination for improved user interaction and experience.
Implementation Method 1
The one or more structured light emitters illuminate the eye with a structured light pattern (e.g., a pattern of infrared radiation). The one or more cameras capture images (e.g., capture images as video stream) of the eye illuminated with the structured light pattern. Parallax between the emitter and the camera may result in distortion in an image (e.g., a frame of a video stream) of the illumination pattern.
Data Source
AI summary
Disclosed is a system and method for tracking a user's eye using structured light. The structured light system is calibrated by training a model of surface of the user's eye. A structured light emitter projects a structured light pattern (e.g., infrared structured light) onto a portion of the surface of the eye. From the viewpoint of a camera, the illumination pattern appears distorted. Based on the distortion of the illumination pattern in the captured image, the eye tracking system can determine the shape of the portion of the user's eye that the structured light is incident upon. By comparing the determined shape of the portion of the user's eye to the model, the orientation of the eye may be determined. The eye tracking system or elements thereof may be part of a head-mounted display, e.g., as part of a virtual reality system.