Event-Based Pupil Detection via Uniform Background Signal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Eye-tracking technologies using event-based sensors for pupil detection are power-hungry due to significant processing requirements from large amounts of data, as most pixels around the pupil experience significant brightness changes, leading to numerous events that demand substantial processing power.
Innovation Solution
The method involves directing a first light source towards the eye and applying a uniform background signal to the event-based sensor, setting a detection threshold so that the pupil's brightness change triggers events while the rest of the eye's change does not, thereby optimizing power consumption by reducing unnecessary processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If event-based sensor is used for dark pupil detection with intermittent illumination, then detection speed is improved, but processing power consumption increases due to large amount of events from pixels around the pupil experiencing brightness changes
Solution Approach 1:
The patent applies local quality by creating different illumination conditions for different regions of the eye. The side illumination creates a local brightness change in the pupil region that is distinct from the surrounding areas, causing only pixels in the pupil region to generate events. This localizes the processing to the relevant area, reducing overall processing power consumption while maintaining fast detection speed.
Solution Approach 2:
The patent inverts the conventional bright pupil detection approach by using dark pupil detection with side illumination. Instead of illuminating the eye from the front to make the pupil bright, it illuminates from the side to make the pupil dark relative to its surroundings, creating contrast through the pupil's light absorption property. This inversion enables event-based detection to focus only on the pupil region.
2Power
If frame-based camera is used for pupil detection, then processing power is reduced compared to continuous sensing, but detection speed is limited by frame rate and power consumption increases with higher frame rates
Solution Approach 1:
The patent replaces the mechanical frame-based camera system with an event-based sensor system. Instead of capturing images at fixed frame rates and processing entire frames, the event-based sensor continuously detects only the necessary brightness changes as events. This substitution eliminates the frame rate limitation while reducing processing power consumption by processing only relevant changes rather than entire frames.
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 reduces power consumption by ensuring the pupil's brightness change is more significant than the rest of the eye's, lowering the number of events that require processing and enhancing the efficiency of pupil detection.
Implementation Method 1
observing the eye with the event-based sensor
Data Source
AI summary
A method for detecting a pupil of an eye with an event-based vision sensor, comprises the steps of directing a first light source towards the eye; observing the eye with the event-based vision sensor; and intermittently submitting the event-based vision sensor to a uniform background signal.