Fingerprint Sensor Pixel Defect Correction via Dispersion Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fingerprint recognition accuracy is compromised by contamination and defective pixels in fingerprint images, leading to reduced recognition rates, especially in mobile devices where security relies heavily on precise user authentication.
Innovation Solution
A method and device that detect defective pixels in fingerprint images by analyzing their dispersion value over time, interpolating pixel values using adjacent pixels, and correcting the image through techniques like linear or spline interpolation to enhance recognition accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional fingerprint recognition is used without correction, then the process is simple, but the recognition rate decreases due to defective pixels and contamination
Solution Approach 1:
The patent performs preliminary detection of defective pixels and contamination in the fingerprint image before recognition. By identifying problematic areas in advance and correcting them through interpolation and filtering operations, the system ensures high-quality input for the recognition algorithm, thereby improving recognition reliability without adding significant complexity to the overall process
Solution Approach 2:
The system performs self-correction by automatically detecting defective pixels and contamination, then generating corrected pixel values through interpolation from surrounding valid pixels. This self-service mechanism eliminates the need for manual intervention or external correction systems, maintaining device complexity at acceptable levels while significantly improving recognition rates
2Reliability
If small foreign materials appear in fingerprint image, then the image contains useful information, but the recognition rate is markedly reduced
Solution Approach 1:
The patent extracts and identifies contaminated regions in the fingerprint image by analyzing pixel dispersion values and comparing them against threshold criteria. Once contamination is detected, the system separates the contaminated pixels from the valid fingerprint data, allowing the recognition algorithm to process only the clean portions of the image, thereby maintaining high recognition rates even when foreign materials are present
Solution Approach 2:
The system converts the harmful effect of contamination into a beneficial detection opportunity. By analyzing deviations in pixel values caused by foreign materials, the system identifies and flags these regions, then uses interpolation to reconstruct the underlying fingerprint pattern that would have been obscured. This transforms the contamination from a recognition obstacle into a detectable and correctable artifact
3Reliability
If pixel values are corrected using interpolation, then defective pixels are fixed, but processing time increases
Solution Approach 1:
The patent applies interpolation correction only to pixels that are identified as defective through dispersion analysis, rather than processing the entire image uniformly. This partial action approach focuses computational resources only where needed, correcting problematic pixels while leaving valid pixels unchanged, thereby minimizing processing time overhead while maintaining high image quality
Data Source
AI summary
Provided are a method of recognizing a fingerprint and a device including the same. The method of recognizing a fingerprint includes obtaining a fingerprint image from a fingerprint sensor, determining whether a pixel in the fingerprint image is defective based on a dispersion value of the pixel with respect to time, in response to determining that the pixel is defective, generating an interpolated pixel value of the pixel by performing interpolation on a first pixel value of the pixel based on a second pixel value of another pixel different from the pixel, and correcting the fingerprint image based on the interpolated pixel value.