Driver Eye-Opening Analysis for Accurate Drowsiness Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining driver drowsiness, such as those based on steering movements, eye status analysis, blinking frequency, and pupil diameter, suffer from inaccuracies due to variations in individual responses and environmental factors, particularly affecting ethnic differences and lighting conditions.
Innovation Solution
A method that determines drowsiness by calculating the average eye-opening measure and duration between specific limiting values, using a camera system to analyze the distance between eyelids, and correlates this with established drowsiness scales to provide accurate drowsiness assessments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the PERCLOS algorithm is used to determine drowsiness by summing eye closure times, then the drowsiness detection can be implemented with simple calculation, but the accuracy is insufficient because it cannot distinguish between drowsy drivers with low blinking frequency and non-drowsy drivers with high blinking frequency
Solution Approach 1:
The patent changes the parameter being measured from simple eye closure time summation to the duration of the opening process of the eye. By focusing on the opening duration rather than closure time, the system achieves better accuracy in detecting drowsiness while maintaining relatively simple calculation requirements. The opening duration is determined by measuring when the eye-opening measure is between lower and upper limiting values, providing a more reliable indicator of drowsiness state.
2Measurement precision
If blinking frequency analysis is used to determine drowsiness, then the method can account for individual variations in eye closure behavior, but the accuracy remains insufficient because many drowsy drivers show increased blink duration rather than increased blinking frequency
Solution Approach 1:
Instead of analyzing the closing phase of blinking as in traditional methods, the patent inverts the approach by analyzing the opening phase of the eye. By measuring the duration during which the eye-opening measure is between lower and upper limiting values (the opening process), the system captures a more reliable indicator of drowsiness that works across different ethnic groups and lighting conditions, without requiring complex blinking frequency analysis.
3Measurement precision
If pupil diameter analysis is used to determine drowsiness, then continuous monitoring is possible, but the accuracy is poor because both drowsy and non-drowsy drivers have different pupil diameters during the day, at night, and under blinding effects that do not correlate with drowsiness level
Solution Approach 1:
The patent extracts the specific measurement of eye-opening measure (distance between upper and lower eyelid) from the broader pupil diameter analysis. By focusing solely on the eyelid distance and determining the duration of the opening process between defined limiting values, the system eliminates the confounding factors associated with pupil diameter variations due to lighting conditions, time of day, and individual differences, achieving more reliable drowsiness detection across diverse conditions.
4Measurement precision
If steering movement analysis is used to determine drowsiness, then non-contact detection is achieved, but the accuracy is very inaccurate for a significant fraction of drivers
Solution Approach 1:
The patent introduces an intermediary measurement - the eye-opening measure (distance between eyelids) - that serves as a more reliable mediator between driver state and detection system. Unlike steering movements which can vary widely among individual drivers, the eye-opening measure provides a consistent physiological indicator that correlates with drowsiness across different drivers, ethnic groups, and operating conditions, significantly improving detection accuracy while maintaining non-contact operation.
Data Source
AI summary
The invention relates to a method for determining the drowsiness of a driver of a motor vehicle. An average eye-opening measure of an open eye of the driver is determined. The eye-opening measure corresponds to a distance between an upper and a lower eyelid of the eye. A lower limit value and an upper limit value are also determined. The lower and upper limit values each correspond to an eye-opening measure less than the determined average maximum eye-opening measure. An average duration of an opening process of the eye of the driver is determined. The eye-opening measure is between the lower and upper limit values during the opening process. The drowsiness of the driver is determined on the basis of the determined average duration of the opening process.