Eyelid Outline Detection for Vehicle Drowsiness Assessment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing drowsiness detection systems in vehicles often falsely detect the edge of the pupil as an eyelid outline, leading to incorrect drowsiness estimation.
Innovation Solution
An image processing device that detects eyelid outlines from face images by determining whether the change in position of detected candidate lines matches normal eyelid movement during blinking, using predetermined conditions to validate the eyelid outlines and prevent false detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If eyelid outline detection is performed using edge detection from face images, then eyelid position can be identified, but false detection occurs where pupil edges are mistakenly detected as eyelid outlines
Solution Approach 1:
The patent applies preliminary action by detecting blink events before performing eyelid outline detection. The system first identifies when a blink occurs, then uses this temporal information to guide the subsequent eyelid detection process, ensuring that detection only occurs during appropriate physiological states rather than attempting to detect eyelid outlines continuously which leads to false positives from pupil edges
Solution Approach 2:
The patent applies dynamics by using the temporal characteristics of blinking to dynamically adjust the detection process. Instead of static edge detection that continuously identifies eyelid outlines, the system dynamically activates detection based on detected blink events, matching the natural temporal pattern of eyelid movement and avoiding false detections during non-blink periods
Data Source
AI summary
An object of the present invention is to reduce false detection of an eyelid from a face image. According to the present invention, it is determined whether the amount of the change in the position of an eyelid outline candidate line during blinking matches the normal movement of an eyelid. When it is determined that the amount of the change in the position of the eyelid outline candidate line does not match the normal movement of the eyelid during blinking, the eyelid outline candidate line is not set as an eyelid outline. Therefore, it is possible to reduce false detection of the eyelid from the face image.