Eye-Gaze Door Access Control for Accurate User Intent Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current access control systems inaccurately determine user intent due to reliance on proximity and face orientation, leading to unnecessary door openings, increased security risks, and resource wastage.
Innovation Solution
An eye-gaze based system for door opening that uses long-range gaze estimation without user calibration, employing a camera with infrared illuminators and machine learning algorithms to categorize gaze as 'on' or 'not on' the door, integrating head orientation and distance measurements for accurate intent detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If proximity-based detection is used to determine user intent, then the system is simple to implement, but the accuracy of intent detection deteriorates leading to false door openings
Solution Approach 1:
The system transitions from 2D face orientation detection to 3D gaze direction estimation by incorporating depth information from infrared structured light. This adds a depth dimension to the detection, enabling more accurate determination of whether the user is looking at the door by calculating the angle between the gaze vector and the door normal vector in three-dimensional space.
Solution Approach 2:
The patent introduces an intermediary calibration process that establishes a mapping between eye position and gaze direction. This intermediary calibration model acts as a bridge between the raw image data and the final gaze direction calculation, enabling accurate intent detection without requiring complex real-time computational models.
2Speed
If face orientation detection is used to determine user intent, then the detection process is fast, but the accuracy deteriorates because users may face the door without intending to open it
Solution Approach 1:
The system replaces mechanical head orientation detection with optical gaze tracking. By using infrared illuminators and structured light patterns reflected from the cornea, the system can detect eye gaze direction optically without requiring mechanical sensors or complex image processing of facial features, maintaining speed while improving accuracy.
Solution Approach 2:
The patent changes the detection parameter from coarse face orientation to fine-grained gaze direction. By measuring the position of corneal reflections and pupil center, the system can determine the precise direction of the user's gaze, distinguishing between merely facing the door and actually looking at it with intent to open.
3Ease of operation
If calibration-free long-range gaze estimation is implemented, then ease of operation improves, but device complexity increases due to infrared illuminators and machine learning models
Solution Approach 1:
The system performs preliminary calibration during manufacturing by projecting structured light patterns and capturing reference images to establish the mapping between eye position and gaze direction. This preliminary action stores calibration data that eliminates the need for user calibration, simplifying operation while distributing the complexity to the manufacturing process rather than user interaction.
Solution Approach 2:
The patent uses structured light patterns as a reference copy that is projected onto the user's face. By capturing the deformation of this known light pattern through corneal reflection, the system can accurately determine gaze direction without requiring complex real-time modeling, effectively using the projected pattern as a reference copy for measurement.
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
Accurately determines user intent for door opening or secure area access, reducing security risks and resource wastage by minimizing false openings.
Implementation Method 1
employing a camera with infrared illuminators
Data Source
AI summary
Systems and techniques may be used for allowing a person access to a secure area, opening a door, or the like. A technique may include detecting a person within a particular distance of a secure area, capturing, using a camera, a set of images of at least a facial feature of the person, preprocessing the set of images to identify the facial feature in at least one image of the set of images, and identifying, using processing circuitry, a gaze status of the person based on the identified facial feature in the at least one image. The technique may include determining, using a trained machine learning model, an intent of the person to access the secure area based on the gaze status, and based on determining the intent, granting access to the secure area.