• Sapna Sharma
• Dr Rashmi Priya

Today's interconnected world, the online environment is crowded with data from billions of sensors, smartphones, smart watches, and other Internet of Things devices, cloud-based services, which leads to hacking threats for everyone. Securing our confidential information with proper authentication is a challenge. In this paper we propose a low cost reliable biometric system based on finger vein patterns. However, finger vein images are easily affected by uneven illumination, humidity ,finger pressure and collection posture, so quality of images gets reduced, resulting in lower performance of authentication method. To address this problem, we worked on the finger vein ROI identification method and image enhancement method by removing the noise using linear and non-linear filters. We have developed an effective CNN model in order to extract features for feature matching with enrolled images. The experiment is performed on MATLAB R2021b using the SDUMLA-HMT database. Analysis of accuracy scores and execution time, exhibit the outstanding performance of Resnet101 with the 97.64% accuracy.

CNN; Feature extraction; Finger vein; ROI; PSNR; Accuracy

"CNN based Feature Extraction for Personal Authentication System using Finger Vein ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.9, Issue 10, page no.c809-c817, October-2022, Available :http://www.jetir.org/papers/JETIR2210307.pdf

"CNN based Feature Extraction for Personal Authentication System using Finger Vein ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.9, Issue 10, page no. ppc809-c817, October-2022, Available at : http://www.jetir.org/papers/JETIR2210307.pdf