Low Resolution Fingerprint Image Verification using CNN Filter and LSTM Classifier

Abstract

A biometric system is an evolving technology that is used in various fields like forensics, secured area and security system. One of the main biometric system is fingerprint recognition system. The reduced rate of performance of fingerprint verification system is due to many reasons such as displacement of finger during scanning, moisture on scanner, etc. The result and accuracy of fingerprint recognition depends on the presence of valid minutiae. According to literature several Fingerprint Recognition System are designed that uses various techniques in order to reduce false detection rate and to enhance the performance of the system. A comparative study of different recognition technique along with their limitations is also summarized and optimum approach is proposed which may enhance the performance of the system. This research work is focused on designing of fingerprint verification/classification including feature extraction methods and learning models for proper classification to label different fingerprints. In order to gain above mentioned objectives, FVC2002 dataset is taken for training and testing. In this dataset there are approx. 72 images which are used for testing purpose. In this dataset there are some blur, distorted as well as partial images also which are considered for recognition. Convolution Neural Network (CNN) and Long Short Term Memory (LSTM) is used for recognition of fingerprint. The result analysis shows approx. 3% enhancement over existing work.

A Robust Automatic Fingerprint Recognition System Using Multi-Connection Hopfield Neural Network

“…This method added these wavelet features to the minutiae features for detection. Tamrakar and Gupta [24] proposed a novel technique for fingerprint recognition in which the CNN network is used for enhancement and feature extraction of fingerprint images. This technique also uses an LSTM network that is trained by a combined set of minutiae and CNN features extracted from input fingerprints.…”

Novel Deep Learning Technique to Improve Resolution of Low-Quality Finger Print Image for Bigdata Applications

“…Since their network is deeper, it is computationally expensive. Ayushi Tamrakar and Neetesh Gupta [17] proposed SR model based on convolutional neural network (CNN) and long shortterm memory (LSTM). In their model, they first applied CNN for both feature extraction and enhancement, thereafter LSTM applied on this feature map to classify images based on ridges in output image so that personal identification is employed using these ridges information.…”