Single Camera Driver Drowsiness Detection via Motion Vector Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional drowsiness driving alarm systems face challenges in accurately detecting a driver's drowsiness due to factors like internal illuminance, external weather conditions, and the use of glasses, and are costly and computationally intensive, especially with stereo camera configurations.
Innovation Solution
A driver state sensing system utilizing a single-channel camera to extract motion vectors from driver images, comparing them with pre-stored drowsiness vectors to determine drowsiness and trigger warnings, minimizing computational and cost burdens while enabling real-time monitoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a stereo camera configuration is used to detect driver motion, then the accuracy of driver state detection is improved, but the cost and computational requirements increase significantly
Solution Approach 1:
The patent extracts only the essential information needed for driver state detection by using a single camera to capture 2D images, rather than using a full stereo camera system. The system extracts motion vectors from sequential 2D images to determine driver drowsiness, eliminating the need for complex 3D image extraction while maintaining detection capability.
Solution Approach 2:
The patent creates a simplified copy of the stereo camera functionality by using a single camera to capture sequential images and computing motion vectors between frames. This approach replicates the motion detection capability of stereo cameras without requiring multiple simultaneous image sensors, thereby reducing hardware complexity and cost.
2Measurement precision
If biological signal detection devices are mounted to detect driver state, then the accuracy of drowsiness detection is improved, but the cost increases
Solution Approach 1:
The patent replaces biological signal detection devices (such as heart rate monitors or brain wave sensors) with an optical imaging system. By using a single camera to capture visual information and analyzing motion vectors in the images, the system substitutes mechanical/optical detection for biological signal detection, thereby maintaining detection accuracy while reducing system cost and complexity.
3Device complexity
If a single camera is used to photograph driver images, then the cost and computational requirements are reduced, but the ability to accurately detect driver motion may be compromised
Solution Approach 1:
The patent performs preliminary actions by capturing sequential images at different time points and pre-calculating motion vectors between these frames. By establishing a temporal sequence of images and computing motion information in advance, the system compensates for the limitations of a single camera and enables accurate driver state detection without requiring complex real-time 3D processing.
Data Source
AI summary
Disclosed are a driver state sensing system, a driver state sensing method, and a vehicle including the same. The driver state sensing system includes an image photographing unit composed of a single channel to photograph images, a controller that extracts at least one shape point from a driver image sensed in the images and compares a motion vector calculated based on the amount of change of the shape point with at least one drowsiness vector stored in advance to determine whether a driver is in a drowsiness state, and a warning unit to warn when it is determined that the driver is in a drowsiness state.