Face Authentication Device Mirror Region Lateral Inversion Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing face authentication technologies using mirrors face challenges such as increased storage capacity and potential for erroneous authentication due to the need to register multiple images and lack of effective mirror detection and correction, especially in environments where camera placement is restricted.
Innovation Solution
A face authentication system that detects the mirror region in input images, applies lateral inversion processing, and compares the processed image with a single registered reference face image to perform authentication, reducing storage requirements and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple face images (front image, reflected front image, reflected side image) are registered with the database to improve authentication accuracy using mirrors, then authentication accuracy is improved, but database storage capacity increases threefold
Solution Approach 1:
The patent applies lateral inversion processing to convert reflected front images and reflected side images into their inverted forms, then registers only the original front image and the laterally-inverted front image in the database. This inversion approach reduces storage requirements while maintaining authentication accuracy by enabling the system to recognize faces regardless of whether they appear in reflected or non-reflected states.
Solution Approach 2:
Instead of storing three separate face images (front, reflected front, reflected side), the patent creates a laterally-inverted copy of the front image and stores only these two images in the database. This copying strategy eliminates redundant storage of reflected images while preserving the ability to authenticate faces captured in mirror reflections.
2Measurement precision
If reflected front image and reflected side image are registered to improve authentication accuracy, then authentication accuracy is improved, but it becomes difficult to distinguish between real images and reflected images when the person does not stand still
Solution Approach 1:
The patent registers the laterally-inverted front image in the database instead of storing reflected images. During authentication, the system laterally inverts the captured image and compares it with the registered laterally-inverted reference image. This approach maintains reliability because the lateral inversion operation is deterministic and consistent, allowing accurate distinction between real and reflected images even when the person is moving.
3Adaptability or versatility
If camera is installed in restricted environments (narrow floor, passage, or ceiling vicinity) to meet installation requirements, then installation flexibility is improved, but the angle, distance, and height cannot be set as recommended, reducing authentication accuracy
Solution Approach 1:
The patent compensates for suboptimal camera angles and positions by applying lateral inversion processing to the captured images. This allows the system to maintain high authentication accuracy even when cameras are installed in restricted environments where recommended angles and distances cannot be achieved, as the lateral inversion corrects for the distorted perspectives.
Solution Approach 2:
The patent introduces lateral inversion processing as an intermediary step between image capture and authentication. This processing mediator transforms images captured from restricted positions into a form that can be accurately compared with registered references, bridging the gap between installation constraints and authentication requirements.
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 efficient face authentication using mirror-reflected images without increasing database storage, enhancing authentication accuracy and flexibility in camera placement, while reducing the risk of erroneous authentication.
Implementation Method 1
a mirror image processing circuitry configured to detect the mirror region from the input image and to subject a detected mirror region to lateral inversion processing to produce a processed image
Data Source
AI summary
Face authentication can easily be performed based on an input image including a mirror region reflected by a mirror without increasing a storage capacity of a database. With the database, for each face of a person to be authenticated, only one real front image is to be registered as a reference face image. A mirror image processing circuitry detects the mirror region from the input image and to subject a detected mirror region to lateral inversion processing to produce a processed image. A face detection processing circuitry detects a face image from the processed image to generate a detected face image. A face authentication processing circuitry compares the detected face image with the reference face image, to thereby perform face authentication.