Fingerprint Minutiae Encoding Vectors for Faster Cross-Database Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for encoding dactylograms require images, which may not be available in some databases, leading to interoperability issues and inefficiencies, and existing graph neural network methods struggle with identifying a single dactylogram among multiple candidates.
Innovation Solution
A method that encodes dactylograms using a graph neural network to convert minutiae coordinates into a fixed-size vector, utilizing a projection and aggregation process to create a representative encoding vector, independent of image availability and format.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If dactylograms are stored as images, then visual inspection and analysis are facilitated, but memory space requirements increase and computing time during identification operations increases
Solution Approach 1:
The patent extracts only the essential features (minutiae) from the complete dactylogram image, storing only these discrete points with their coordinates and characteristics. This extraction process eliminates the need to store and process entire images while retaining the critical information needed for identification, thus reducing computing time without completely sacrificing analytical capability.
Solution Approach 2:
The patent creates a simplified representation (copy) of the dactylogram in the form of minutiae points rather than storing the original image. This copied form maintains the essential identification features while requiring significantly less storage space and enabling faster processing during identification operations.
2Loss of time
If dactylograms are classified into various classes based on morphological features, then identification time is reduced, but the complexity of the classification system increases
Solution Approach 1:
The patent segments the dactylogram into discrete minutiae points (terminations, bifurcations, lakes, bridges) rather than attempting to classify the entire image as a whole. This segmentation allows for simpler, more automated classification based on individual feature characteristics, reducing both identification time and system complexity compared to traditional morphological classification methods.
3Quantity of substance
If only minutiae coordinates are stored instead of images, then memory space is reduced and interoperability is improved, but the ability to perform visual analysis and quality assessment is lost
Solution Approach 1:
The patent extracts and stores only the critical quantitative information from the dactylogram (minutiae coordinates, types, and relationships), discarding the redundant visual data. This extraction achieves significant memory space reduction while retaining sufficient information for automated identification operations.
4Productivity
If automated identification systems are implemented, then processing speed increases, but the requirement for image quality and format standardization increases
Solution Approach 1:
The patent transforms dactylogram data from image format to a standardized parameter-based representation (minutiae coordinates and characteristics). This parameter change enables automated processing at high speeds while improving adaptability across different databases and systems, as the standardized format can be universally processed without requiring image quality standardization.
Data Source
AI summary
Computer-implemented method for encoding a list of minutiae of a dactylogram, said method taking, as input data, the coordinates associated with each minutia from a list of minutiae of a dactylogram, and providing, as output datum, a fixed-size encoding vector which is representative of the list of the minutiae of said dactylogram, the method comprising the following steps:(a) concatenating the coordinates of each minutia from the list of minutiae in the form of a source matrix of dimension;(b) projecting the source matrix of dimension into a space of a dimension greater than the dimension of said source matrix using a projection model previously trained to form a projected matrix of dimension;(c) inferring an inference matrix of dimension by applying, to the intermediate matrix, a previously trained graph neural network;(d) aggregating the values of the inference matrix into a fixed-size vector using a previously defined aggregation model, said fixed-size vector being the fixed-size encoding vector which is representative of the list of the minutiae of the dactylogram.