Fingerprint Reconstruction Using Shared Image Segments
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fingerprint verification systems using small area sensors face challenges in user experience and accuracy due to the need for multiple impressions with precise overlap, leading to inefficiencies and increased chances of false rejections, especially when processing images with limited computing and memory resources.
Innovation Solution
A system and method for multiplicating pattern images by identifying a reconstruction set of impressions that match or correlate with each other, producing initial and additional reconstructions from different combinations of these impressions, reducing user participation and improving robustness without additional user input.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple impressions are collected to cover the entire fingerprint pattern using a small area sensor, then the coverage and completeness of the trusted image improves, but the complexity of operation and user experience deteriorates due to difficulty in achieving precise overlap
Solution Approach 1:
The system automatically determines whether collected impressions provide sufficient unique information for reconstruction without requiring user judgment or manual intervention. The processor autonomously evaluates the overlap and uniqueness of each impression, eliminating the need for users to understand or control the technical parameters of fingerprint capture.
Solution Approach 2:
The system provides real-time feedback to the user about the registration progress and quality. The processor evaluates each collected impression and communicates back to the user whether sufficient unique information has been obtained, allowing the user to understand the status without needing to understand the underlying technical complexity.
2Reliability
If multiple impressions are collected to improve pattern coverage, then the quality of reconstructed pattern improves, but the processing time and computational resources increase
Solution Approach 1:
The system collects more impressions than the minimum theoretically required for reconstruction. By gathering excess impressions and then selectively using only those providing unique information, the system ensures sufficient coverage while the processor efficiently discards redundant data, balancing quality requirements with processing efficiency.
Solution Approach 2:
The processor extracts and identifies only the unique information portions from each collected impression, separating useful data from redundant overlapping regions. This extraction process efficiently processes multiple impressions by focusing computational resources only on the novel information each image contributes.
3Loss of information
If multiple impressions are collected to ensure sufficient coverage, then the completeness of fingerprint data improves, but the chance of false rejection increases due to variations in reconstruction
Solution Approach 1:
The system performs preliminary evaluation of collected impressions to determine whether sufficient unique information has been obtained before completing the reconstruction. This preliminary assessment ensures that the registration process only proceeds when adequate data is available, preventing incomplete reconstructions that would lead to false rejections during verification.
Solution Approach 2:
The system dynamically adjusts the criteria for accepting impressions based on the accumulated unique information. As more diverse impressions are collected, the system modifies its evaluation parameters to ensure sufficient coverage of the entire fingerprint pattern, adapting the registration requirements to the actual data collected rather than using fixed thresholds.
Data Source
AI summary
A system, method, and computer program product for multiplicating pattern images used as a trusted resource while reducing user participation, such as, for example, a producing N number (N>1) of reconstructed pattern images using shared images common to at least two constructed pattern images.