Deep neural networks based wrist print region segmentation and classification

In recent years, biometric recognition based systems have become widespread. One of these is wrist-based recognition systems. In this study, wrist print based recognition system was developed by using near infrared (NIR) camera. Totally 220 NIR camera images taken from 10 for each both hands of 11 p...

Full description

Saved in:
Bibliographic Details
Main Authors: Kerim Kürşat Çevik, H. Erdinç Kocer
Format: Article
Language:English
Published: Kyrgyz Turkish Manas University 2021-06-01
Series:MANAS: Journal of Engineering
Subjects:
Online Access:https://dergipark.org.tr/en/download/article-file/1488991
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849407852280545280
author Kerim Kürşat Çevik
H. Erdinç Kocer
author_facet Kerim Kürşat Çevik
H. Erdinç Kocer
author_sort Kerim Kürşat Çevik
collection DOAJ
description In recent years, biometric recognition based systems have become widespread. One of these is wrist-based recognition systems. In this study, wrist print based recognition system was developed by using near infrared (NIR) camera. Totally 220 NIR camera images taken from 10 for each both hands of 11 people. The obtained data set is allocated 70% (154 images) for training and 30% (66 images) for testing. The wrist regions are labeled on the training set images. Data sets were created with two different labeling methods. In the first data set, only the wrist regions were labeled and it was aimed to segment the wrist region from the image. In the second data set, the wrist images were labeled according to 22 classes and these classes were tried to be predicted. The labeled data was trained with YOLOV2 architecture supported by ResNet50 one of the deep neural network models. The trained model was tested with the remaining 30% of the data set. In the test process, the wrist region was determined in the NIR images with the trained model. As a results of the study, it was seen that the wrist regions were correctly detected in all first data set test images and the mean value of obtained similarity rates was 95.26%. In the test results of the second dataset, 92.43% classification success was obtained. Therefore, it can be said that the deep learning architectures ResNet and YOLO are effective in the segmentation of the wrist region.
format Article
id doaj-art-43862dee041a4b6ea8d97bb97a1cbf5f
institution Kabale University
issn 1694-7398
language English
publishDate 2021-06-01
publisher Kyrgyz Turkish Manas University
record_format Article
series MANAS: Journal of Engineering
spelling doaj-art-43862dee041a4b6ea8d97bb97a1cbf5f2025-08-20T03:35:55ZengKyrgyz Turkish Manas UniversityMANAS: Journal of Engineering1694-73982021-06-0191303610.51354/mjen.8539711437Deep neural networks based wrist print region segmentation and classificationKerim Kürşat Çevik0https://orcid.org/0000-0002-2921-506XH. Erdinç Kocer1https://orcid.org/0000-0002-0799-2140AKDENİZ ÜNİVERSİTESİSelçuk UniversityIn recent years, biometric recognition based systems have become widespread. One of these is wrist-based recognition systems. In this study, wrist print based recognition system was developed by using near infrared (NIR) camera. Totally 220 NIR camera images taken from 10 for each both hands of 11 people. The obtained data set is allocated 70% (154 images) for training and 30% (66 images) for testing. The wrist regions are labeled on the training set images. Data sets were created with two different labeling methods. In the first data set, only the wrist regions were labeled and it was aimed to segment the wrist region from the image. In the second data set, the wrist images were labeled according to 22 classes and these classes were tried to be predicted. The labeled data was trained with YOLOV2 architecture supported by ResNet50 one of the deep neural network models. The trained model was tested with the remaining 30% of the data set. In the test process, the wrist region was determined in the NIR images with the trained model. As a results of the study, it was seen that the wrist regions were correctly detected in all first data set test images and the mean value of obtained similarity rates was 95.26%. In the test results of the second dataset, 92.43% classification success was obtained. Therefore, it can be said that the deep learning architectures ResNet and YOLO are effective in the segmentation of the wrist region.https://dergipark.org.tr/en/download/article-file/1488991wrist print recognitiondeep neural networksnear-infrared camerayolo
spellingShingle Kerim Kürşat Çevik
H. Erdinç Kocer
Deep neural networks based wrist print region segmentation and classification
MANAS: Journal of Engineering
wrist print recognition
deep neural networks
near-infrared camera
yolo
title Deep neural networks based wrist print region segmentation and classification
title_full Deep neural networks based wrist print region segmentation and classification
title_fullStr Deep neural networks based wrist print region segmentation and classification
title_full_unstemmed Deep neural networks based wrist print region segmentation and classification
title_short Deep neural networks based wrist print region segmentation and classification
title_sort deep neural networks based wrist print region segmentation and classification
topic wrist print recognition
deep neural networks
near-infrared camera
yolo
url https://dergipark.org.tr/en/download/article-file/1488991
work_keys_str_mv AT kerimkursatcevik deepneuralnetworksbasedwristprintregionsegmentationandclassification
AT herdinckocer deepneuralnetworksbasedwristprintregionsegmentationandclassification