Noncontact Biometric Identification Using Dynamic Hand Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Noncontact biometric identification devices face challenges in maintaining high accuracy and usability due to fluctuations in hand posture and background interference during palm vein authentication, as existing methods fail to systematically provide enrollment data with sufficient variations and are not effective in multidimensional scenarios.
Innovation Solution
A noncontact biometric identification device that uses a sensor to detect feature data at multiple positions while the hand moves, linking this data to position and posture information, and selects enrollment data based on hyperspheres in a feature space defined by position and posture coordinates, ensuring maximum radii and non-overlap, to optimize image selection and authentication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a user's hand is held on the sensor to maintain optimal posture, then identification accuracy is improved, but usability is degraded
Solution Approach 1:
The system dynamically tracks the hand posture and continuously captures vein images while the hand moves, rather than requiring a static optimal position. The enrollment data selection unit dynamically selects the most appropriate image from the continuous stream based on real-time posture evaluation, allowing the system to maintain accuracy while accommodating natural hand movement.
Solution Approach 2:
The system changes the evaluation criteria from requiring a single optimal static posture to evaluating multiple dynamic postures. By establishing evaluation indexes that assess image quality across various hand positions and angles, the system can select enrollment data that represents the best match regardless of the user's natural hand movement patterns.
2Ease of operation
If vein images are continuously captured while trailing the moving hand, then usability is improved, but identification accuracy may degrade without proper image selection
Solution Approach 1:
The system implements feedback through evaluation indexes that continuously assess the quality of captured vein images based on hand posture. The enrollment data selection unit uses this feedback to automatically select the most appropriate image from the continuous stream, ensuring that only high-quality images meeting predefined criteria are used for enrollment, thus maintaining accuracy while allowing continuous capture.
Solution Approach 2:
The system performs self-service by automatically evaluating and selecting enrollment data from the continuous stream of captured images without requiring user intervention. The enrollment data selection unit autonomously determines which images meet the evaluation criteria and selects them as enrollment data, eliminating the need for manual image selection while maintaining high identification accuracy.
3Ease of manufacture
If a single static image is captured during registration, then enrollment process is simple, but accuracy degrades when non-optimum image is registered
Solution Approach 1:
Instead of capturing a single image, the system captures an excessive number of vein images continuously while the hand moves, then selects the most appropriate one based on evaluation indexes. This partial action approach captures more images than necessary but ensures that at least one high-quality image meeting the evaluation criteria is selected, thereby maintaining accuracy without significantly increasing user burden.
Solution Approach 2:
The system performs self-service by automatically evaluating and selecting enrollment data from the continuous stream of captured images without requiring user intervention. The enrollment data selection unit autonomously determines which images meet the evaluation criteria and selects them as enrollment data, eliminating the need for manual image selection while maintaining high identification accuracy.
4Adaptability or versatility
If different operations are performed between enrollment and comparison, then flexibility is improved, but enrollment data appropriateness cannot be ensured
Solution Approach 1:
The system implements multi-functionality by using the same evaluation indexes for both enrollment and comparison operations. The enrollment data selection unit and the collation result determination unit both rely on identical evaluation criteria to assess image quality and select appropriate data, ensuring consistency across different operations while maintaining operational flexibility.
Solution Approach 2:
The system changes the evaluation criteria from requiring a single optimal static posture to evaluating multiple dynamic postures. By establishing evaluation indexes that assess image quality across various hand positions and angles, the system can select enrollment data that represents the best match regardless of the user's natural hand movement patterns, ensuring appropriateness across different operational contexts.
Data Source
AI summary
A biometric identification device acquires enrollment data in such a manner as to improve usability while maintaining identification accuracy. The noncontact biometric identification device includes: a sensor that detects feature data including a feature of a part of a living body at a plurality of positions in a range defined in advance in a real space when the part of the living body moves in the range to acquire enrollment data; a processor that arranges a specified number of hyperspheres in a comparison range along a line defined by a locus of the living body in the real space such that the hyperspheres run over the comparison range, do not overlap one another, and have maximum radii, and then selects the feature data detected at one of the positions in the real space closest to the position corresponding to the center of one of the hyperspheres as the enrollment data.