Driver Visual Recognition Judgment via Eye Fluctuation Patterns
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional safety-drive assistance devices face challenges in accurately determining whether a driver recognizes important objects during driving, especially with low detection accuracy of the visual-line direction, leading to inefficient warnings and increased costs for high-accuracy systems.
Innovation Solution
A safety-drive assistance device that judges visual recognition based on the fluctuation pattern of a driver's eye direction, eliminating the need for high-accuracy visual-line direction detection and reducing costs, by using a combination of cameras and sensors to detect moving objects and track eye movement patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a device uses high-accuracy visual-line direction detection (accuracy under 0.17 degrees) to judge driver recognition, then the judgment accuracy improves, but the device cost increases significantly
Solution Approach 1:
The patent replaces expensive high-accuracy visual-line detection devices with cheaper alternatives (infrared camera + image processing) that achieve sufficient accuracy through software algorithms rather than hardware precision, reducing device cost while maintaining functional effectiveness
Solution Approach 2:
The patent changes the detection parameter from requiring sub-0.17 degree visual-line accuracy to accepting around 1 degree accuracy compensated by fluctuation pattern analysis, transforming the measurement approach to achieve comparable judgment reliability with lower precision requirements
2Reliability
If a device warns a driver only when detecting overlooked objects to ensure safety, then collision prevention improves, but the device may give unnecessary warnings when the driver already knows the dangerous state
Solution Approach 1:
The patent uses the driver's eye movement fluctuation patterns as feedback to infer awareness state, continuously monitoring whether the driver is paying attention to relevant objects and adjusting warning generation based on this inferred awareness level
Solution Approach 2:
The patent replaces the mechanical assumption that detection = need for warning with a biological/behavioral model using eye movement patterns to infer driver awareness, substituting physical detection accuracy with behavioral analysis
3Device complexity
If a device judges visual recognition using visual-line direction overlapping frequency, then the judgment method simplifies, but the detection accuracy must be under 0.17 degrees which is difficult to achieve
Solution Approach 1:
The patent adds a temporal dimension to the spatial visual-line detection problem by analyzing fluctuations over time, transforming a 2D spatial overlapping problem into a 3D spatiotemporal pattern recognition problem that tolerates larger spatial uncertainties
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables effective judgment of a driver's visual recognition of critical objects, even with low detection accuracy of the visual-line direction, thereby providing timely and necessary warnings for potential collisions or pedestrian contact.
Implementation Method 1
there is a method of reflecting a purkinje image of which detecting accuracy is 0.3 to 0.4 using a near infrared ray camera
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention is to provide a safety-drive assistance device that can judges whether a driver recognizes an object to which a driver should pay an attention or not even if an accuracy of detecting a visual-line direction is almost equal to one degree. A traffic environment detecting unit 11 detects an object such as an automobile, a pedestrian, a road mark and a traffic signal. A watched target judging unit 12 judges a watched target among the objects, to which a driver pays attention. A visual-line direction detecting unit 13 detects a visual-line direction of a driver. A visual recognition judging unit 14 judges whether a driver visually recognizes the watched target or not, based on the watched target and a fluctuation pattern of a direction of either or both eyes of the driver. A non-safety recognition judging unit 15 judges whether a driver recognizes non-safety or not, based on the result of visual recognition judged by the visual recognition judging unit 14. A presentation unit 16 informs the content of non safety, based on the result of judging non safety judged by the non-safety recognition judging unit 15.