Small-Area Biometric Sensor Using Multispectral Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fingerprint-based biometric sensors face challenges in reducing size without compromising functionality or usability, as their performance degrades with smaller sizes, and existing methods like mosaiking require precise user interaction and larger enrollment areas.
Innovation Solution
The use of multispectral skin texture analysis with small-area sensors that measure locally consistent optical properties, such as spatiospectral imaging, to enable biometric identification without the need for large enrollment areas, using spatial, spectral, and textural descriptors combined with classification methodologies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If the sensor size is reduced, then the device area and cost are reduced, but the biometric performance degrades
Solution Approach 1:
The patent changes the measurement parameters from conventional single-wavelength or narrow-band imaging to multispectral imaging across multiple wavelengths. This parameter change enables the extraction of locally consistent biometric features from smaller sensor areas, resolving the contradiction between reduced sensor size and maintained biometric performance
Solution Approach 2:
The patent adds the spectral dimension to the traditional spatial imaging approach. By incorporating wavelength as an additional dimension and analyzing spatio-spectral features, the system can achieve reliable biometric identification with smaller sensor areas, as the spectral information provides additional discriminative power that compensates for reduced spatial sampling
2Area of stationary object
If mosaiking or swipe methods are used to reduce sensor size, then the sensor area is reduced, but user operation complexity increases
Solution Approach 1:
The patent enables the system to automatically acquire sufficient biometric data from a single finger placement without requiring user learning or precise positioning. The multispectral imaging approach extracts multiple features simultaneously from the captured image, allowing the system to serve itself by obtaining adequate enrollment data in one simple operation
3Loss of information
If conventional fingerprint imaging is used, then sufficient biometric data is obtained, but the sensor size is large
Solution Approach 1:
The patent extracts only the essential biometric features needed for identification from the finger image, rather than requiring capture of the entire fingerprint pattern. By focusing on locally consistent features and using multispectral analysis, the system extracts sufficient biometric information from smaller sensor areas, eliminating the need for large-area conventional fingerprint imaging
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 allows for a compact, highly usable biometric sensor that maintains functionality across various applications by capturing locally consistent features, reducing errors due to skin movement and improving usability and reliability.
Implementation Method 1
receiving light scattered from the small-area purported skin site
Data Source
AI summary
Systems, devices, methods, and software are described for biometric sensors that permit a reduction in the size of the sensing area without significant reduction in biometric functionality of the sensor. A skin site of an individual is illuminated, and light scattered from the skin site is received. An image of a locally consistent feature of the skin site is formed from the received light. The locally consistent feature is analyzed to perform a biometric function.