Fingerprint Recognition Loop with Image Restoration for Adverse Conditions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fingerprint recognition rates are often reduced due to sweat, stains, or dry fingers, affecting user experience in mobile terminals.
Innovation Solution
A fingerprint recognition method and system that collects images in loops, uses image restoration algorithms like total variation methods to repair damaged images, and matches them with pre-stored templates, with a mechanism to track collection attempts and switch between different templates based on similarity thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional fingerprint recognition methods are used, then the system is simple and fast, but the recognition rate drops when sweat, stain, or dry conditions affect the fingerprint
Solution Approach 1:
The patent applies preliminary action by pre-storing multiple fingerprint templates (including normal and problematic fingerprints like sweat, stain, and dry conditions) before recognition. When a fingerprint image is captured, the system compares it against these pre-stored templates to find the best match, significantly improving recognition rates for problematic fingerprints without adding complex real-time processing
Solution Approach 2:
The patent changes the parameter of fingerprint template diversity by storing multiple templates with different quality characteristics (normal, sweaty, stained, dry). This parameter change allows the system to adapt to various finger conditions by selecting the most appropriate template for matching, thereby improving reliability without requiring complex adaptive algorithms
2Reliability
If multiple fingerprint templates are pre-stored for the same finger, then the recognition rate for problematic fingerprints improves, but the storage requirement and system complexity increase
Solution Approach 1:
The patent applies partial action by storing only the essential multiple templates needed for common problematic conditions (sweat, stain, dry) rather than all possible variations. This selective approach improves recognition for the most frequent issues while keeping storage requirements manageable
Solution Approach 2:
The patent optimizes the parameter of template quantity by determining the optimal number of templates to store based on statistical analysis of fingerprint conditions. This ensures sufficient templates are stored to handle problematic cases while avoiding excessive storage consumption
Data Source
Figure 1
Figure 2
AI summary
A fingerprint recognition method capable of improving a fingerprint recognition rate includes: Step 1, collecting a fingerprint image, and recording the number of collecting times, and then proceeding to Step 2; Step 2, matching the fingerprint image with pre-stored fingerprint templates for recognition, where a current matching process for recognition ends if matching succeed, otherwise, the method proceeds to Step 3; Step 3, restoring the fingerprint image, and matching the restored fingerprint image with the pre-stored fingerprint templates for recognition, where the current matching process for recognition ends if matching succeed, otherwise, the method proceeds to Step 4; Step 4, comparing the number of collecting times with a preset number of times, where the current matching process for recognition ends if the number of collecting times reaches the preset number of times otherwise, the method returns back to Step 1. The method performs matching and recognition by collecting fingerprint images in loops and performs further matching and recognition by restoring an unrecognized fingerprint image, which may significantly improve the fingerprint recognition success rate when sweat or stain exists on fingers or the fingers are dry.