Facial Expression Analysis for Driver Sleepiness Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for determining sleepiness in drivers, such as analyzing eye movements, are unreliable due to disturbances caused by environmental factors and eye-related indications that vary across different driving conditions, making it difficult to accurately assess sleepiness levels.
Innovation Solution
A system that uses facial expression analysis, including measurements like the distance between mouth corners, eyebrow-eye distance, head tilt angle, and other facial feature points, to determine sleepiness levels by capturing and processing facial image data, allowing for accurate assessment without direct electrode attachment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If eye movement analysis is used to determine sleepiness, then the determination can be made from readily observable features, but the reliability is reduced due to environmental disturbances and variations in eye behavior across different driving conditions
Solution Approach 1:
The invention segments the facial region into multiple distinct feature points (eyes, eyebrows, nose, mouth, chin) and analyzes the relationships between them. By dividing the face into measurable segments with specific coordinate relationships, the system achieves reliable sleepiness determination through multiple independent measurements rather than relying on a single eye movement metric.
Solution Approach 2:
The facial feature point analysis system serves multiple functions: it can detect eye opening, eyebrow position, nose movement, mouth changes, and chin position all from the same set of facial landmarks. This multi-functional approach using universal facial feature points provides robust sleepiness determination that works across various driving conditions without requiring separate detection mechanisms.
2Reliability
If multiple facial expression information parameters are measured to improve sleepiness determination accuracy, then the reliability increases, but the device complexity increases due to multiple measurement requirements
Solution Approach 1:
The invention merges multiple facial feature detections (eyes, eyebrows, nose, mouth, chin) into a unified coordinate system. By combining all these measurements and analyzing their relative positions and movements together, the system achieves high reliability in sleepiness determination while using a single integrated detection apparatus rather than multiple separate devices.
Solution Approach 2:
The system creates a digital model (copy) of the driver's face by detecting and storing the positions of multiple facial feature points. This facial map serves as a simplified representation that captures essential geometric relationships without requiring complex physical measurement devices, thereby reducing device complexity while maintaining measurement accuracy.
3Ease of operation
If facial feature points are detected using image processing, then non-contact measurement is achieved, but the measurement precision may be affected by image quality and processing complexity
Solution Approach 1:
The invention applies local quality by focusing image processing resources on specific facial regions and feature points rather than attempting to analyze the entire face uniformly. Each facial landmark (eyes, eyebrows, nose, mouth, chin) is detected with specialized attention to its local geometric characteristics, improving measurement precision while maintaining efficient non-contact operation.
Data Source
AI summary
A sleep prevention system captures a facial image of a driver, determines a sleepiness level from the facial image and operates warning devices including neck air conditioner, seatbelt vibrator, and brake controller if necessary based on the sleepiness determination. A sleepiness determination device determines sleepiness from facial expression information such as distances between corners of a mouth, distance between an eyebrow and eye, tilt angle of a head, and other facial feature distances. The facial distances are calculated from captured images and from reference information gather during wakefulness. The sleepiness degree is determined based on the determination results including combinations thereof.