Drowsiness Detection via Dynamic Heartbeat Normalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing drowsiness detecting devices face challenges in accurately detecting drowsiness due to fluctuations in heartbeat feature values over time, especially for the same individual, leading to inconsistent results.
Innovation Solution
A drowsiness detecting device comprising a feature extractor, normalizer, detecting rule storage unit, and drowsiness detector that normalizes heartbeat feature values using a dynamic normalization coefficient, updated based on confidence levels, to improve detection accuracy across time and individual variations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If heartbeat feature values are used directly for drowsiness detection, then the detection process is simple, but the detection accuracy deteriorates due to fluctuations in feature values over time and individual differences
Solution Approach 1:
The patent applies dynamics by implementing a dynamic normalization coefficient that changes over time based on the user's heartbeat data. The system continuously updates the normalization coefficient using moving average calculations, allowing the detection threshold to adapt to temporal variations in the user's physiological state rather than using a fixed threshold.
Solution Approach 2:
The patent changes the parameter of the normalization coefficient from a fixed value to a dynamically updated value. By calculating the moving average of heartbeat intervals and using this to normalize the feature values, the system transforms the detection parameter to account for individual differences and temporal variations, thereby improving detection accuracy.
2Reliability
If a fixed detection threshold is used, then the detection rule is simple, but the detection reliability deteriorates due to individual differences and temporal variations in heartbeat features
Solution Approach 1:
The patent transforms the fixed detection threshold into a dynamic normalization coefficient that adapts to each user's characteristics. By calculating the moving average of heartbeat intervals and using this as a normalization basis, the system creates a personalized detection threshold that improves reliability across different individuals and time periods.
Solution Approach 2:
The system performs self-adaptation by automatically calculating the normalization coefficient from the user's own heartbeat data without requiring external calibration or manual adjustment. The moving average calculation enables the system to self-adjust to the user's physiological characteristics over time.
Data Source
AI summary
A drowsiness detecting device includes a feature extractor, a detecting rule storage unit, a normalizer, and a drowsiness detector. The feature extractor extracts a feature value on heartbeats of a user based on intervals between the heartbeats. The detecting rule storage unit retains a detecting rule for drowsiness detecting. The normalizer updates a normalization coefficient. The drowsiness detector detects drowsiness of the user based on the feature value, the detecting rule, and the normalization coefficient. The normalizer updates the normalization coefficient based on the feature value when the drowsiness detector does not detect the drowsiness of the user.