Dense U-Net Deep Learning Technique for Semantic Iris Segmentation
K. Vijayakeerthi
, A. Asmitha yugasre , S. Swathi , S. SriVidhya , V. Neela Priyadharshini
Iris segmentation, Unet, DenseNet, Dense Unet, Deep learning, Accuracy.
The process of iris segmentation is crucial to iris identification. To maintain the accuracy of an iris identification system based on iris segmentation, errors existing in the present stage are limited. It works badly because of a non-ideal environment such as ambient lack of user involvement and light noise. Under these non-ideal circumstances, current segmentation techniques based on local gestures are unable to acknowledge the iris bounds, and the mistakes caused at the segmentation step are passed to all future stages, resulting in reducing dependability and accuracy. This research presents a Dense U-Net-based iris segmentation technique to overcome the issues with existing algorithms in complicated scenes and cross- device applications. Iris is segmented by merging a dense network with a traditional U-net, which uses narrower and fewer parameters and information, which makes training deep networks easier. The obtained results were close to the floor truths, the accuracy was also excellent, and our proposed model performs better in classification.
"Dense U-Net Deep Learning Technique for Semantic Iris Segmentation", TIJER - TIJER - INTERNATIONAL RESEARCH JOURNAL (www.TIJER.org), ISSN:2349-9249, Vol.11, Issue 2, page no.377-383, February-2024, Available :https://tijer.org/TIJER/papers/TIJERB001059.pdf
Volume 11
Issue 2,
February-2024
Pages : 377-383
Paper Reg. ID: TIJER_151064
Published Paper Id: TIJERB001059
Downloads: 000297
Research Area: Science and Technology
Country: Thiruvallur, TamilNadu, India
ISSN: 2349-9249 | IMPACT FACTOR: 8.57 Calculated By Google Scholar | ESTD YEAR: 2014
An International Scholarly Open Access Journal, Peer-Reviewed, Refereed Journal Impact Factor 8.57 Calculate by Google Scholar and Semantic Scholar | AI-Powered Research Tool, Multidisciplinary, Monthly, Multilanguage Journal Indexing in All Major Database & Metadata, Citation Generator
Publisher: TIJER (IJ Publication) Janvi Wave
