Scribd
Upload
103 views

Development of open-CV Framework For Minutiae Extraction and Matching of Fingerprints

The document discusses developing an open-CV framework for minutiae extraction and matching of fingerprints. It aims to address limitations in current fingerprint recognition systems, mainly longer computation time for fingerprint matching. The proposed approach extracts minutiae features from fingerprint images and creates an OpenCV structure for minutiae extraction and matching, with the goal of making fingerprint recognition more optimized, faster and efficient.

Uploaded by

Sunitha Manam
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
103 views

Development of open-CV Framework For Minutiae Extraction and Matching of Fingerprints

The document discusses developing an open-CV framework for minutiae extraction and matching of fingerprints. It aims to address limitations in current fingerprint recognition systems, mainly longer computation time for fingerprint matching. The proposed approach extracts minutiae features from fingerprint images and creates an OpenCV structure for minutiae extraction and matching, with the goal of making fingerprint recognition more optimized, faster and efficient.

Uploaded by

Sunitha Manam
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 5

Development of open-CV framework for minutiae

Extraction and matching of fingerprints

Problem statement: Fingerprint Recognition is a vital method in biometric


identification and verification of human beings in various domains like Security,
Digital Forensics, Internet of Things (IoT), and many more. Each individual
human is having distinct fingerprint pattern than others, hence it is one of the most
prominent and widely used method to distinguish individuals. Many research
studies and solutions have been developed in biometric domain since a decade,
which influences now in making the process of fingerprint recognition more
optimized, faster and efficient. However, present fingerprint
acquisition/recognition systems have some limitations, mainly longer computation
time for fingerprint matching and evaluating the results. This paper presents a
procedure for fingerprint matching that takes into account minutiae features in
finger print images and the process of creating an OpenCV structure for minutiae
extraction and matching of fingerprints.

Scope: Fingerprint Recognition is a vital method in biometric identification and


verification of human beings in various domains like Security, Digital Forensics,
Internet of Things (IoT), and many more. Each individual human is having distinct
fingerprint pattern than others, hence it is one of the most prominent and widely
used method to distinguish individuals. So, it’s very important for us to distinguish
individuals for security point of view.
Who said? What? Existing solution:

Authors: J. Wu, Z. Liu, J. Li, C. Gu, M. Si, and F. Tan

Title:Biometric methods to speed up fingerprint processing


using OCT: A survey
Fingerprint recognition is a crucial security technique with a steady growing usage
for the identification and verification of people. However, current fingerprint
acquisition systems have bound disadvantages that embody the necessities of
physical contact with the acquisition device, and also the presence of undesirable
artefacts, like scars on the surface of the fingerprint. There are various sectors or
domains where Biometric authentication plays a vital role, these areas include
Internet of Things (IoT), Smartphones, Healthcare, and many more. This paper
evaluates the accuracy of a whole framework for the capturing of uninjured,
undistorted fingerprints from underneath the skin's surface using Optical
Coherence Tomography (OCT), the extraction and conversion of the submersed
knowledge into a usable fingerprint and also the matching of such fingerprints. The
main objective of this paper is to provide the study of the biometric algorithms,
OCT-based methods and their relative comparisons.

Authors: J. Wu, Z. Liu, J. Li, C. Gu, M. Si, and F. Tan


Title: Partial Fingerprint Identification Through
Checkerboard Sampling Method Using ANN
This paper presents a straightforward method of fingerprint identification
considering partial fingerprints through neural network system. The fingerprint
templates constitute the largest data in the biometric field. By considering partial
fingerprint samples, the amount of data in the database shall be decreased and this
will consequently lead to a faster processing for fingerprint identification. To
address the problem on database capacity, sampling the full fingerprint is done
through checkerboard algorithm. A standard 8x8 checkerboard is adopted wherein
selected blocks are considered as samples for Artificial Neural Network (ANN)
training. This will reduce the memory size of a fingerprint template to 50%. Also,
25% of the whole fingerprint is done to further reduce the memory size. The
fingerprint images underwent the usual image enhancement by FFT and histogram
equalization. Then, image binarization is done to be able to use pattern
identification tool in an artificial neural network system available in MATLAB.

Authors: J. Wu, Z. Liu, J. Li, C. Gu, M. Si, and F. Tan


Title: Fingerprint Recognition Using Global and Local
Structures
Biometrics is one of the biggest tendencies in human identification. The fingerprint
is the most widely used biometric. However considering the automatic fingerprint
recognition a completely solved problem is a common mistake. The global level
structures consist of many ridges to form some specific shape like arch, loop, and
whorl. Local level structures are called minutiae, which further classified as either
endpoints or bifurcations. Either of which can be used to identify the fingerprint,
our approach uses both methods.

Authors: J. Wu, Z. Liu, J. Li, C. Gu, M. Si, and F. Tan


Title:Speeding up and enhancing a large-scale fingerprint
identification system on GPU
Fingerprint identification is one of the most common biometric feature problems
which is used in many applications. Although state-of-the-art algorithms are very
accurate, the need for fast processing a big database of millions of fingerprints is
highly demanding. In this paper, we propose to adapt the fingerprint matching
algorithm based on Minutia Cylinder-Code on Graphics Processing Units to speed
up the matching. Another contribution of this paper is to add a consolidation stage
after matching to enhance the precision. The experimental results on a GTX-680
and K40 tesla devices with standard data-sets prove that the proposed algorithm
can be comparable with the state-of-the-art approach, and is suitable for a real-time
identification application.

Methodologies and Drawback of Existing solution:


In existing system, a method of checkerboard sampling algorithm using ANN for
partial fingerprint identification. Authors claim that this work can save memory
space in the database. It also introduces an encryption technique for the whole
fingerprint. Sun Bei presented a method that accomplishes a decent classification
performance of imperfect or partially corrupted fingerprint images. The method
involves two steps. Initially a skeleton model of corrupted fingerprint in created
using preprocessing techniques. Secondly, a finger print verification system is
created that identifies the core point of the corrupted fingerprint and recreates or
models the trivial features called minutiae. Thirdly, a neural network architecture is
developed for classification of fingerprints.

How it is related to problem statement: In proposed system we are


going to extract the minutiae features in fingerprint images and the process of
creating an OpenCV structure for minutiae extraction and matching of fingerprints.
But if we compare it existing solution where they are using different methods like
checkerboard sampling algorithm using ANN for partial fingerprint identification
or creating the skeleton model of corrupted images and after fingerprint matching
using deep learning models.

How the problem is going to solved: In making the process of


fingerprint recognition more optimized, faster and efficient. However, present
fingerprint acquisition/recognition systems have some limitations, mainly longer
computation time for fingerprint matching and evaluating the results. This paper
presents a procedure for fingerprint matching that takes into account minutiae
features in finger print images and the process of creating an OpenCV structure for
minutiae extraction and matching of fingerprints.

You might also like