Fingerprint Sensor Calibration for Silkscreen Logo Interference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fingerprint recognition technologies face inaccuracies due to the embedding of silkscreened logo patterns in fingerprint modules, causing dielectric constant differences between the silkscreen and glue layers, which result in a clear logo image overlapping with the fingerprint image, affecting recognition accuracy.
Innovation Solution
A method and apparatus for calibrating fingerprint images by obtaining prestored information about non-fingerprint images, specifically the gain coefficients of fingerprint sensor units, and using this information to remove the non-fingerprint image part from the collected fingerprint image, improving recognition accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a logo pattern is silkscreened on the fingerprint module cover, then the device branding is enhanced, but the fingerprint recognition accuracy deteriorates due to dielectric constant differences causing logo image overlap
Solution Approach 1:
The patent applies preliminary action by pre-calibrating the fingerprint sensor to identify and compensate for the logo pattern interference before actual fingerprint recognition occurs. The system stores calibration data representing the logo image and uses it to subtract the logo pattern from captured fingerprint images, thereby eliminating the harmful effect of the silkscreened logo on recognition accuracy while preserving the branding benefit
Solution Approach 2:
The patent converts the harmful logo pattern into a beneficial calibration reference. By treating the logo image as a known quantity that can be measured and stored, the system transforms the interference source into a useful subtraction mask that, when applied during image processing, enhances fingerprint recognition accuracy by removing the logo overlay
2Reliability
If the silkscreen layer and glue layer have different dielectric constants, then the logo pattern becomes visible in fingerprint images, but this creates calibration errors that degrade image quality
Solution Approach 1:
The patent implements feedback by using the visible logo pattern as a reference signal that informs the calibration process. The system captures images of the logo pattern, processes them to create a calibration mask, and feeds this information back into the image processing pipeline to continuously compensate for dielectric constant differences, thereby maintaining image quality despite the physical layer variations
3Adaptability or versatility
If fingerprint sensor units are used to detect fingerprint images, then biometric authentication is enabled, but the presence of embedded logo patterns causes false recognition results
Solution Approach 1:
The patent applies the extraction principle by separating the logo pattern component from the fingerprint image through calibration-based subtraction. The system extracts the logo image characteristics during calibration, creates a separate mask representation, and then removes this extracted logo component from the captured fingerprint images, leaving only the pure fingerprint data for accurate biometric authentication
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The calibration method effectively removes non-fingerprint image parts from collected fingerprint images, enhancing the accuracy and recognition effectiveness of fingerprint images by accounting for the dielectric constant differences and logo image interference.
Implementation Method 1
Dielectric constants of the silkscreen layer and the glue layer are not completely the same, resulting in a significant difference between a fingerprint sensor unit corresponding to the protrusion and other fingerprint sensor units
Data Source
AI summary
A method and an apparatus for calibrating a fingerprint image, and a terminal, are provided. The method for calibrating a fingerprint image includes steps of: obtaining prestored information about a non-fingerprint image; and collecting a fingerprint image, and calibrating the fingerprint image according to the information about the non-fingerprint image. The method can remove a non-fingerprint image part from a collected fingerprint image and improve accuracy of the collected fingerprint image, thereby improving a fingerprint recognition effect.