Corneal Reflex Feature Extraction for Single-Image Biometric Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing biometric detection technologies require multiple images to perform authentication, which increases processing burden and costs, and are not effective with a single image.
Innovation Solution
A processing system that extracts and compares the feature amount of light reflected by a cornea from a single image with reference features from various imaging apparatuses to determine if the image is from a living body, reducing the need for multiple images and complex camera setups.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple images are used for biometric detection based on corneal reflex, then detection accuracy is improved, but processing burden and costs increase
Solution Approach 1:
The patent extracts and utilizes the corneal reflex characteristics from a single image to perform biometric detection. By focusing on the specific feature of corneal light reflection rather than requiring multiple images, the system achieves accurate detection while reducing processing burden. The extraction unit identifies corneal reflex points and uses their positional relationships to determine authentication results.
Solution Approach 2:
The patent performs preliminary extraction of corneal reflex feature amounts from the single image before comparison. The system pre-processes the image to identify and extract characteristic points of corneal reflection, then compares these extracted features against reference data. This preliminary extraction enables accurate detection without requiring multiple images or complex post-processing.
2Reliability
If multiple images are required for biometric detection, then authentication reliability is improved, but the number of imaging operations increases
Solution Approach 1:
The system extracts sufficient authentication information from a single image by focusing on corneal reflex characteristics. The extraction unit identifies specific features such as the position and shape of corneal light reflections, which contain unique biometric information. This extraction approach maintains authentication reliability while reducing the number of imaging operations required.
Solution Approach 2:
The patent changes the detection parameter from requiring multiple images to analyzing specific corneal reflex features within a single image. By shifting focus to the unique optical properties of corneal reflection and their positional relationships, the system achieves reliable authentication with faster imaging speed and fewer operational steps.
3Productivity
If a single image is used for biometric detection, then processing efficiency is improved, but detection accuracy may deteriorate
Solution Approach 1:
The patent applies local quality analysis by focusing on specific regions and features within the single image - particularly the corneal reflex points and their characteristics. Instead of requiring multiple images, the system extracts detailed local features from the cornea area, including the position, shape, and intensity distribution of reflected light. This localized detailed analysis compensates for using only a single image while maintaining high detection accuracy.
Solution Approach 2:
The system adds dimensional analysis by examining the spatial relationships and geometric properties of corneal reflex points within the single image. The comparison unit analyzes the positional relationships between multiple reflex points, their distances from reference points, and their angular arrangements. This multi-dimensional feature extraction from a single two-dimensional image enables accurate detection without requiring multiple images.
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 biometric detection based on a single image, reducing processing burden and costs, while effectively distinguishing between living and non-living subjects.
Implementation Method 1
extracts a feature amount of light reflected by a cornea of a person from an image generated by a first imaging apparatus
Data Source
AI summary
The example embodiments provides a processing system including an extraction unit that extracts a feature amount of light reflected by a cornea of a person from an image generated by a first imaging apparatus, a comparison unit that specifies an imaging apparatus corresponding to the extracted feature amount of the light based on the extracted feature amount of the light and a reference feature amount indicating a feature of light emitted by each of a plurality of imaging apparatuses at a time of imaging, and a biometric detection unit that performs biometric detection based on a result of whether or not the first imaging apparatus matches the specified imaging apparatus.