IoT Based Cryptolock with Face Recognition Using AES-256 Bit Algorithm

Smart door lock system is one of the renowned applications of IoT. In an IoT based system, the user is transmitting his confidential data over the network which may contain passwords, images and other information. This paper proposes a cryptolock, which can monitor invalid passage with face recognition based on Convolutional Neural Network (CNN). The proposed method shows good recognition accuracy for CNN based approach. The security and privacy of the biometric system is ensured by encrypting the user records using Advanced Encryption Standards(AES). Captured face image of the recipient will be transmitted over the network after hashing(SHA-256) and encryption(AES-256), further authentication can be done using decryption(AES-256) of the hash vector by the user to unlock the door. Both hashing and encryption/decryption keys are generated from user claimed password. The objective is to develop a password based cryptographically shielded highly secure door lock system to resolve the valuables security and information protection issues. Google Drive can be used as the platform for remote access and it gives a complete sense of security.