Polarized Infrared Eye Detection for Driver Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing face image processing technologies face challenges in accurately detecting the positions and states of eyes, especially under varying brightness conditions and external light influences, which can lead to false recognition and unreliable detection of driver attention or drowsiness.
Innovation Solution
A face image processing apparatus utilizing polarized light, with a first polarizer emitting infrared light and a second polarizer capturing images, generates difference images to detect hyperbolic or cross-shaped patterns present only in infrared images, compensating for external light influences and accurately determining eye positions and states.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional face image processing is used to detect eye positions, then the detection can be performed under normal lighting conditions, but the detection accuracy deteriorates under varying brightness conditions and external light influences
Solution Approach 1:
The patent introduces polarized light as an intermediary to detect eye positions. By using a light source that emits polarized light and a polarization filter in the image capturing device, the system creates a controlled optical environment that isolates eye reflections from external light sources. The polarized light interacts with the corneal surface to produce characteristic reflection patterns that can be detected regardless of ambient lighting conditions, thus resolving the contradiction between detection reliability and external light interference.
Solution Approach 2:
The patent changes the optical parameters of the detection system by utilizing polarized light with specific orientation. The light source emits polarized light at a predetermined angle, and the polarization filter is oriented to detect specific reflection patterns from the cornea. This parameter change in light polarization allows the system to distinguish eye reflections from other facial features and external light sources, maintaining detection accuracy under varying brightness conditions.
2Measurement precision
If conventional image processing methods are used, then the system structure remains simple, but measurement precision of eye positions deteriorates due to false recognition
Solution Approach 1:
The patent employs polarized light as an intermediary to enhance measurement precision. The light source emits polarized light that reflects off the corneal surface, and the polarization filter in the image capturing device detects these reflections. This intermediary optical mechanism provides distinctive reflection patterns from the eyes that are easily distinguishable from other facial features, thereby improving eye position precision without requiring complex image processing algorithms.
Solution Approach 2:
The patent utilizes changes in light polarization state as a visual indicator for eye detection. The polarized light reflects off the cornea with characteristic polarization patterns that appear as distinct bright regions in the captured images. These polarization-based visual changes provide clear, unambiguous signals for eye position detection, improving measurement precision while keeping the system structure relatively simple.
3Object-affected harmful factors
If polarization filtering is applied to capture images, then external light influence is reduced, but the device complexity increases due to additional polarizer components
Solution Approach 1:
The patent implements periodic switching between different polarization filter orientations. The polarization filter alternates between capturing images with different polarization angles, allowing the system to isolate eye reflections from external light sources through differential analysis. This periodic action enables the system to eliminate external light interference by comparing images taken at different polarization states, effectively reducing harmful external light influence.
Solution Approach 2:
The patent segments the image capture process into multiple frames taken with different polarization filter orientations. By dividing the detection process into sequential captures with varying polarization angles, the system can process each frame independently and combine the results to eliminate external light interference. This segmentation approach manages device complexity by breaking down the complex polarization filtering into manageable, sequential steps.
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
This approach enables reliable detection of eye positions and states, ensuring accurate assessment of driver attention and normal driving conditions by compensating for external light and reducing false recognition.
Implementation Method 1
a first polarizer which polarizes light in a first direction and a light emitter, through the first polarizer, which emits infrared light; an image capturing portion including a second polarizer which polarizes light in a second direction perpendicular to the first direction
Data Source
AI summary
A face image processing apparatus includes: a lighting portion including a first polarizer which polarizes light in a first direction and a light emitter which emits, through the first polarizer, infrared light; an image capturing portion including a second polarizer which polarizes light in a second direction perpendicular to the first direction and an image capturing unit which captures images through the second polarizer; and an image processing portion which detects candidates of eyes using a first image captured when the lighting portion emits the infrared light and a second image captured when the lighting portion does not emit the infrared light. The image processing portion determines, as an eye, a candidate having a hyperbolic or cross shaped pattern present in the first image but not present in the second image.