Dorsal Hand Vein Pattern Recognition: A Comparison between Manual and Automatic Segmentation Methods
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Objectives: Various techniques for dorsal hand vein (DHV) pattern extraction have been introduced using small datasetswith poor and inconsistent segmentation. This work compared manual segmentation with our proposed hybrid automaticsegmentation method (HHM) for this classification problem. Methods: Manual segmentation involved selecting a region-ofinterest(ROI) in images from the Bosphorus dataset to generate ground truth data. The HHM combined histogram equalizationand morphological and thresholding-based algorithms to localize veins from hand images. The data were divided intotraining, validation, and testing sets with an 8:1:1 ratio before training AlexNet. We considered three image augmentationstrategies to enlarge our training sets. The best training hyperparameters were found using the manually segmented dataset. Results: We obtained a good test accuracy (91.5%) using the model trained with manually segmented images. The HHMmethod showed slightly inferior performance (76.5%). Considerable improvement was observed in the test accuracy of themodel trained with the inclusion of automatically segmented and augmented images (84%), with low false acceptance andfalse rejection rates (0.00035% and 0.095%, respectively). A comparison with past studies further demonstrated the competitivenessof our technique. Conclusions: Our technique can be feasible for extracting the ROI in DHV images. This strategyprovides higher consistency and greater efficiency than the manual approach.
Abstract
Objectives: Various techniques for dorsal hand vein (DHV) pattern extraction have been introduced using small datasetswith poor and inconsistent segmentation. This work compared manual segmentation with our proposed hybrid automaticsegmentation method (HHM) for this classification problem. Methods: Manual segmentation involved selecting a region-ofinterest(ROI) in images from the Bosphorus dataset to generate ground truth data. The HHM combined histogram equalizationand morphological and thresholding-based algorithms to localize veins from hand images. The data were divided intotraining, validation, and testing sets with an 8:1:1 ratio before training AlexNet. We considered three image augmentationstrategies to enlarge our training sets. The best training hyperparameters were found using the manually segmented dataset. Results: We obtained a good test accuracy (91.5%) using the model trained with manually segmented images. The HHMmethod showed slightly inferior performance (76.5%). Considerable improvement was observed in the test accuracy of themodel trained with the inclusion of automatically segmented and augmented images (84%), with low false acceptance andfalse rejection rates (0.00035% and 0.095%, respectively). A comparison with past studies further demonstrated the competitivenessof our technique. Conclusions: Our technique can be feasible for extracting the ROI in DHV images. This strategyprovides higher consistency and greater efficiency than the manual approach.
참고문헌 (30)
- Chowdhury AM, "Contactless fingerprint recognition using deep learning : a systematic review", J Cybersecur Priv, 2(3), pp.714~730 (2022)[DOI]
- Laghari WA, "Dorsal hand vein identification using transfer learning from AlexNet", Int J Integr Eng, 14(3), pp.111~119 (2022)[DOI]
- Dhieb T, "A novel biometric system for signature verification based on score level fusion approach", Multimed Tools Appl, 81(6), pp.7817~7845 (2022)[DOI]
- Zulfiqar M, "Deep face recognition for biometric authentication", Proceedings of 2019 International Conference on Electrical, Communication, and Computer Engineering (ICECCE), pp.1~6 (2019)
- Nithya AA, "Iris recognition techniques : a literature survey", Int J Appl Eng Res, 10(12), pp.32525~32546 (2015)
- Jia W, "A survey on dorsal hand vein biometrics", Pattern Recognit, 120, pp.108122~ (2021)[DOI]
- Kumar R, "International Conference on Innovative Computing and Communications", pp.1087~1107 (2021), Springer
- Rajalakshmi M, "Palm-dorsal vein pattern authentication using convoluted neural network(CNN)", Int J Pure Appl Math, 116(23), pp.525~532 (2017)
- Raghavendra R, "Hand dorsal vein recognition: Sensor, algorithms and evaluation", Proceedings of 2015 IEEE International Conference on Imaging Systems and Techniques (IST), pp.1~6 (2021)
- Khan MH, "Investigating linear discriminant analysis (LDA) on dorsal hand vein images", Proceedings of the 3rd International Conference on Innovative Computing Technology (INTECH), pp.54~59 (2013)
- Al-johania NA, "Dorsal hand vein recognition by convolutional neural networks : feature learning and transfer learning approaches", Int J Intell Eng Syst, 12(3), pp.178~191 (2019)[DOI]
- Li X, "Biometric recognition", pp.296~306 (2016), Springer
- Mohaghegh M, "Automated biometric identification using dorsal hand images and convolutional neural networks", J Phys Conf Ser, 1880, pp.012014~ (2021)[DOI]
- Wang J, "Quality-specific hand vein recognition system", IEEE Trans Inf Forensic Secur, 12(11), pp.2599~2610 (2017)[DOI]
- Guo Z, "A novel algorithm of dorsal hand vein image segmentation by integrating matched filter and local binary fitting level set model", Proceedings of 2020 7th International Conference on Information Science and Control Engineering (ICISCE), pp.81~85 (2020)
- Chanthamongkol S, "Dorsal hand vein image enhancement for improve recognition rate based on SIFT keypoint matching", Proceedings of the 2nd International Symposium on Computer, Communication, Control and Automation, pp.174~177 (2013)
- Sontakke BM, "Automatic ROI extraction and vein pattern imaging of dorsal hand vein images", Int J Sci Adv Res Technol, 4(3), pp.1678~1683 (2018)
- Manju RA, "Improved method for enhancing dark images based on CLAHE and morphological reconstruction", Procedia Comput Sci, 165, pp.391~398 (2019)[DOI]
- Chen L, "Near-infrared dorsal hand vein image segmentation by local thresholding using grayscale morphology", Proceedings of 2007 1st International Conference on Bioinformatics and Biomedical Engineering, pp.868~871 (2007)
- Sinan S. Mohammed Sheet, "Retinal Disease Identification using Upgraded CLAHE filter and Transfer Convolution Neural Network", ICT Express, 8(1), pp.142~150 (2022), 한국통신학회[DOI]
- Hitam MS, "Mixture contrast limited adaptive histogram equalization for underwater image enhancement", Proceedings of 2013International Conference on Computer Applications Technology (ICCAT), pp.1~5 (2013)
- Chin SW, "Dorsal hand vein authentication system using artificial neural network", Indones J Electr Eng Comput Sci, 21(3), pp.1837~1846 (2021)[DOI]
- Akram MU, "Dorsal hand veins based person identification", Proceedings of 2014 4th International Conference on Image Processing Theory, Tools and Applications (IPTA), pp.1~6 (2014)
- Hofbauer H, "Exploiting superior CNN-based iris segmentation for better recognition accuracy", Pattern Recognit Lett, 120, pp.17~23 (2019)[DOI]
- LeCun Y, "Gradient-based learning applied to document recognition", Proc IEEE, 86(11), pp.2278~2324 (1998)[DOI]
- Lefkovits S, "CNN approaches for dorsal hand vein based identification", Proceedings of International Conference in Central Europe on Computer Graphics, Visualization and Computer Vision (WSCG), pp.51~60 (2019)
- Malik J, "Reference threshold calculation for biometric authentication", Int J Image Graph Signal Process, 2, pp.46~53 (2014)[DOI]
- Wijewickrama R, "Write to know:on the feasibility of wrist motion based user-authentication from handwriting", Proceedings of the 14th ACM Conference on Security and Privacy in Wireless and Mobile Networks, pp.335~346 (2021)
- Picart-Armada S, "Benchmarking network propagation methods for disease gene identification", PLoS Comput Biol, 15(9), pp.e1007276~ (2019)[DOI]
- "Junyulong Technology product center live biometrics" (2021)[URL]