Project Title:
Fingerprint and Signature Recognition System with Deep Learning.
Project Guide:
Project Guide: Y Harshalatha, Asst. Professor, Department. of Electronics and Communication, Siddaganga Institute of Technology, Tumakuru, Karnataka, India.
Project Members :
Amruth Y
Gopinatha B M
Gowtham M P
Kiran M
Project Description:
This project aims at developing a novel system for validation of signature and fingerprint using Convolutional Neural network (CNN). The signature of the individual is captured using webcam, processed and is accepted only if it is from intended person. Parallelly the fingerprint is captured in the form of image using the sensor and is verified.
In this system, a camera captures the image of the signature and an optical sensor captures the fingerprint in the form of an image, both the images are pre-processed and fed to the neural network. The neural network model is trained with 80 percent of the images and remaining 20 percent are used for testing. TensorFlow framework is used to train deep learning algorithm on google collaboratory platform. The model is converted to lite version and it is deployed on raspberry pi processor for real time verification. The necessary action is taken only if the accepted bio metric inputs are authenticated by the system.
Publication:
Amruth, Y., Gopinatha, B.M., Gowtham, M.P., Kiran, M. and Harshalatha, Y., 2020, November. Fingerprint and Signature Authentication System using CNN. In 2020 IEEE International Conference for Innovation in Technology (INOCON) (pp. 1-4). IEEE.
Hardware Description:
Raspberry pi 4 4gb Model B
USB Web camera
R307 optical fingerprint sensor
7 inch Touch screen Display
10000mAh xiomi Power Bank
Software Stack:
Raspbian OS
Python 3
Numpy, Matplotlib, Open-CV, Tensorflow
PyQT5
Gallery
Hardware Model
