Fingerprint Image Selection Using Stability and Sharpness Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fingerprint capture systems on mobile devices face challenges with varying backgrounds and image quality, requiring costly image mosaicking and being inefficient in using all types of image sensors for acquiring high-quality fingerprint images quickly and with low memory usage.
Innovation Solution
A method that acquires and selects high-quality fingerprint images by determining image stability and sharpness, calculating a score for each image, and comparing it to a threshold, with optional updates to improve image quality and reduce acquisition time, using a combination of segmentation, convex polygon identification, and stability updates across image sequences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If image mosaicking algorithms are used to reconstruct the fingerprint, then the fingerprint can be captured with a smaller acquisition surface, but the implementation cost increases
Solution Approach 1:
The patent extracts and selects only the most critical quality attributes (sharpness, stability, lighting) from the captured images rather than attempting to reconstruct the entire fingerprint using complex mosaicking algorithms. This selective approach achieves sufficient fingerprint capture quality without the high computational cost and complexity of image mosaicking.
2Adaptability or versatility
If a color camera is used to capture fingerprint images, then the device can use existing image sensors, but the background variability and image quality consistency deteriorate
Solution Approach 1:
The patent performs preliminary quality assessment of captured images by evaluating sharpness, stability, and lighting conditions before selecting images for authentication. This preliminary filtering action ensures that only images meeting minimum quality thresholds are used, maintaining reliable authentication despite the variability introduced by using general-purpose color cameras.
3Measurement precision
If multiple fingerprint images are captured and processed, then the identification precision can be improved, but the acquisition time and memory usage increase
Solution Approach 1:
The patent implements a feedback mechanism where each captured image is immediately evaluated for quality attributes (sharpness, stability, lighting), and only images meeting the criteria are selected for further processing. This feedback loop allows the system to achieve high identification precision through selective image processing while minimizing acquisition time by avoiding unnecessary processing of low-quality images.
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
A method for selecting a fingerprint image for the identification of an individual is described. The method comprises: a) acquiring (S102) a current image including a fingerprint and segmenting said fingerprint; b) determining (S104) a value representing the stability of said current image; c) determining (S106) a value representing the sharpness of said current image; d) determining (S108) a score, said score being a combination of said value representing stability, said value representing sharpness, and a number of segmented fingerprints; e) selecting (S112) said current image for the identification of said individual if said score is greater than a threshold value and, otherwise, storing said current image in memory as the best image if its score is greater than a best-score value and repeating steps a) to e).