Finger Segmentation via Phase Derivative Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for capturing fingerprint images without contact struggle to precisely locate fingers and fingerprints due to unclear image definitions, especially when swipe speed and positioning vary.
Innovation Solution
A finger segmentation method using a processing unit, acquisition device, and projection device to capture and process images with a test pattern, calculating the unwrapped phase image, deriving the phase along the iso-intensity axis, regionalizing pixels by polarity, and recognizing finger locations through region analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If contact-based fingerprint capture is used, then image clarity is improved, but user convenience and hygiene deteriorate
Solution Approach 1:
The patent replaces contact-based mechanical capture with non-contact optical capture using a camera to photograph fingers as they pass through the device, eliminating physical contact while maintaining identification accuracy
Solution Approach 2:
The patent introduces a controlled environment with standardized lighting and a defined capture zone as an intermediary between the user and the capture device, enabling clear images without direct contact
2Ease of operation
If non-contact fingerprint capture is used, then user convenience is improved, but finger location precision deteriorates
Solution Approach 1:
The patent divides the captured image into multiple processing stages: detecting candidate finger regions, verifying them through characteristic analysis, and precisely locating fingerprints within confirmed finger regions to achieve accurate positioning
Solution Approach 2:
The patent implements a verification mechanism that analyzes finger characteristics (length, width, curvature) and provides feedback to confirm or reject candidate regions, ensuring precise finger location identification
Data Source
Figure 1~3
Figure 4a~4b
Figure 5a~7
AI summary
The invention relates to a finger segmentation method (300) comprising: a capture step (302) during which an acquisition device captures an image of the fingers and of a test pattern; a calculation step (304) during which a processing unit calculates the unwrapped phase image on the basis of the thus-captured image of the test pattern; a derivation step (306) during which, for each pixel of the thus-calculated phase image, the processing unit calculates the derivative of the phase on the iso-intensity axis of the test pattern; a regionalisation step (308) during which the processing unit associates each pixel at which the derivative of the phase has been calculated with a region that gradually groups together all of the neighbouring pixels at which the derivative of the phase has the same polarity; and a recognition step (310) during which the location of each finger is recognized from the analysis of the regions thus determined.