Vehicle Gaze Tracking Using Optical Reflections for Occluded Views
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional direct camera imaging in vehicles faces challenges such as occlusions, direct sunlight, and wide or long distance head and eye positions, which impede accurate gaze tracking for multiple occupants.
Innovation Solution
Utilizing multi-surface optical reflections and deep learning models to process reflected images from multiple cameras, enabling independent eye tracking for multiple occupants via face landmark detection, gaze tracking, and occupant location determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If direct camera imaging is used for eye tracking in vehicles, then the system structure is simple, but gaze tracking accuracy deteriorates due to occlusions, direct sunlight, and wide or long distance head and eye positions
Solution Approach 1:
The patent introduces reflective surfaces (mirrors, glass surfaces) as intermediaries between the cameras and the occupants. Instead of directly imaging the occupants' eyes, the system captures reflections of the eyes in these surfaces. This intermediary approach allows the cameras to be positioned away from direct line-of-sight constraints, overcoming occlusions and lighting issues while maintaining gaze tracking accuracy.
Solution Approach 2:
The system transitions from 2D direct imaging to capturing reflections that provide additional spatial dimensions of information. By utilizing multiple reflective surfaces at different angles and positions, the system gains extra dimensional data about eye position and orientation, improving measurement precision without requiring a single complex camera system.
2Measurement precision
If multiple cameras are deployed to improve gaze tracking accuracy, then measurement precision improves, but power consumption increases
Solution Approach 1:
The patent combines the functionality of multiple cameras into a unified reflection-based imaging system. Instead of having separate camera systems for different viewing angles, the system uses a single or limited number of cameras to capture reflections from multiple surfaces, merging the data from what would otherwise require multiple direct camera views into a single processing stream, thereby reducing power consumption.
Solution Approach 2:
The system creates optical copies (reflections) of the occupants' eyes in reflective surfaces and captures these copies with the cameras. This copying approach allows the system to obtain multiple viewing angles and high-quality eye images without requiring multiple direct camera views, reducing the computational and power burden associated with processing multiple camera feeds simultaneously.
3Adaptability or versatility
If image data from all cameras is processed, then comprehensive eye tracking coverage is achieved, but processing efficiency and power usage deteriorate due to substandard image data
Solution Approach 1:
The system performs preliminary evaluation of captured image data to assess quality metrics such as illumination, focus, and reflection clarity before processing. By filtering out substandard images in advance, the system ensures that only high-quality data proceeds to the gaze tracking algorithm, maintaining comprehensive coverage through quality selection while improving processing efficiency by avoiding wasted computation on poor-quality images.
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
Improves gaze tracking accuracy and efficiency by selecting high-quality image data for processing, reducing power consumption, and enhancing user experience in vehicles with reflective interiors.
Implementation Method 1
Multi-user gaze tracking in a vehicle space through evaluation of imaging sources and optical surface reflections
Data Source
AI summary
Methods, systems, and storage media for performing multi-user gaze tracking in a vehicle space using multi-surface optical reflections are disclosed. Implementations may: acquire face and eye region image data of a plurality of occupants within a field of view of at least one camera associated with a vehicle; evaluate reflected image quality thresholds; locate and match occupants within the vehicle space; and perform eye tracking for multiple occupants independently via reflected multi-view images provided to a deep learning model.