Main Article Content
Smart forgery detection for signatures using deep learning and neural networksis a system that will play an important role in authentication of an individual‟s signature. Hence for various security purposes, the above stated system will be useful for instance at banks, legal documentations, government organizations, investigation purposes, business, automatic data entry etc.This system is trained on Convolutional Neural Network (CNN) that works well with the data set of 300 images. CNN has been implemented in Python using Keras with TensorFlow backend. It will compare the data set images with the image (of signature) provided by the user and hence output will be predicted whether the given image is authentic or forged.This system is expected to yield an accuracy of 80-85%. Also this system is practical and accurate enough to be deployed. Hence the above stated system successfully recognizes and identifies the signature of and individual. The planned system is highly economical and its accuracy can be increased by training this model on more data set.
This work is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International License.
CC Attribution-NoDerivatives 4.0