Driving Assistance Driver-State Detection with Adaptive Face Ranges
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional driving assistance systems struggle to accurately determine the anomalous state of a driver when the driver is wearing an article that covers their face, leading to erroneous determinations or failure to detect anomalous states due to reduced feature points in facial recognition.
Innovation Solution
The system identifies the likelihood of a driver wearing an article based on facial feature points, adjusts the state determination period and anomaly range accordingly, and uses these to determine the anomalous state with enhanced precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the driver wears an article covering the face, then the facial recognition system cannot accurately detect driver state, but the system continues to use fixed determination parameters leading to erroneous results
Solution Approach 1:
The patent applies dynamics by making the determination parameters (anomaly range and state determination period) variable rather than fixed. The system dynamically adjusts these parameters based on the detected degree of likelihood that the driver is wearing an article. When a mask is detected, the anomaly range is widened and the determination period is extended, allowing the system to adapt to the obscured facial features and maintain reliable driver state detection.
2Device complexity
If the system uses fixed anomaly range and determination period, then the system structure is simple, but the determination accuracy decreases when facial features are obscured
Solution Approach 1:
The patent implements parameter changes by modifying the anomaly range and state determination period based on the detected likelihood of worn articles. The system calculates different anomaly ranges and determination periods corresponding to different degrees of likelihood (e.g., high, medium, low probability of mask wearing). This allows the system to maintain high determination accuracy across various driving conditions without requiring a completely complex system architecture.
3Measurement precision
If the system adjusts determination parameters based on worn article detection, then determination accuracy improves, but the processing time and computational load increase
Solution Approach 1:
The system applies preliminary action by pre-calculating and storing multiple sets of determination parameters (anomaly ranges and determination periods) corresponding to different degrees of likelihood. When a worn article is detected, the system quickly selects the appropriate pre-prepared parameters based on the detected likelihood level, rather than calculating optimal parameters in real-time. This reduces computational load and determination time while maintaining high accuracy.
Data Source
AI summary
A driving assistance apparatus has: an identifying section that identifies a degree of likelihood that a driver of a vehicle has a worn article on a face on the basis of one or more feature points of the face extracted from captured images; a deciding section that decides a state determination period during which it is determined whether or not a state of the driver is an anomalous state, and an anomaly range representing an orientation of the face observed in a case where the state of the driver is an anomalous state, the state determination period and the anomaly range corresponding to the degree of likelihood; and a determining section that determines whether or not the state of the driver is an anomalous state by determining whether or not an orientation of the face based on the one or more feature points is included in the anomaly range.