Fingerprint Matching via Multidimensional Ridge Feature Projection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fingerprint matching systems face challenges in accurately aligning and matching fingerprints due to intra-class variation, noise, and inconsistent feature extraction, particularly when dealing with low-quality images and elastic deformation, which leads to high false match and false non-match rates.
Innovation Solution
The method involves constructing a canonical framework for fingerprint images by justifying and partitioning them into regions, measuring ridge orientation and separation, and projecting these values into a multidimensional coordinate system to normalize and enhance representation distances, allowing for nonlinear alignment and elastic deformation correction, and incorporating sweat pore features for improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional fingerprint matching algorithms use minutiae information and ridge features for identification, then personal identification can be achieved by searching database templates, but intra-class variation, noise, and elastic deformation lead to high false match and false non-match rates
Solution Approach 1:
The fingerprint image is divided into multiple local regions or blocks, and ridge orientation and separation measurements are performed independently in each region. This segmentation allows the system to handle local variations and deformations more effectively, improving matching accuracy while reducing the impact of noise and elastic deformation.
Solution Approach 2:
The patent transforms fingerprint features into a multidimensional coordinate system by combining ridge orientation, ridge separation, and their derivatives. This dimensional expansion creates a more robust feature space that better distinguishes between genuine matches and false matches, reducing both false match and false non-match rates.
2Measurement precision
If fingerprint images are normalized into a common framework by measuring ridge orientation and separation in multiple regions, then intra-class variation is reduced and matching accuracy is enhanced, but the complexity of the processing system increases
Solution Approach 1:
By dividing the fingerprint image into multiple regions and performing measurements independently in each region, the system achieves better normalization and reduced intra-class variation. This segmented approach manages complexity by breaking down the processing task into smaller, more manageable units that can be handled systematically.
Solution Approach 2:
The patent measures multiple parameters (ridge orientation, ridge separation, and their derivatives) in each region and transforms them into a multidimensional coordinate system. This parameter expansion improves matching accuracy by capturing more discriminative features, while the systematic transformation process manages the complexity through consistent mathematical operations.
3Reliability
If the representation distance between fingerprint images in the multidimensional coordinate system is enhanced, then matching reliability is improved, but the computational requirements and processing time increase
Solution Approach 1:
The system performs preprocessing steps including image justification, partitioning into regions, and measurement of ridge features before the actual matching process. By preparing the data in advance and organizing it into a standardized multidimensional format, the system improves matching reliability while reducing the computational burden during the actual matching operation.
Solution Approach 2:
Transforming features into a multidimensional coordinate system enhances the representation distance between different fingerprint images, improving matching reliability. The structured transformation process, while computationally intensive, is performed once during preprocessing, making subsequent matching operations more efficient and reliable.
Data Source
AI summary
Exemplary embodiments of method and apparatus for processing the images of fingerprints can be provided. For example, aligned images can be subjected to a tessellation process, whereas each image can be partitioned into a number of regions. Within each region at least one parameter associated with the ridges can be measured and stored. Such exemplary parameter can include, e.g., the prevailing ridge orientation, the average ridge separation and the phase of the ridges. The data can be projected and stored in a multidimensional coordinate system, whereas the representations of any two data can be separated by an amount corresponding to the dissimilarity of these data.