Offline Handwritten Signatures Based Multifactor Authentication in Cloud Computing using Deep CNN Model
- Vol.6(22), Jul-Dec
- Tamil Nadu i-manager's 2019
- 13-25p.
Cloud Security is an important factor that influences the adoption of cloud applications into bank domains. Many researchers proposed secure authentication mechanisms based on the traditional factors, biometric factors, captcha and certificates etc. This paper proposes a biometric handwritten signature recognition using Deep Convolution Neural Networks (DCNN). The proposed model uses signature as a biometric factor to verify the authenticity of the users along with traditional credentials. The extraction of the features are performed using DeepCNN model in the registration and verification process. The practical setup is done through NIVIDIA DGX environment using Python keras and tensor flow as backend. An experimental result shows 99% of accuracy and validation accuracy.