Authentication that Combines rPPG Information with Face Detection on the Blockchain

Blockchain is a decentralized distributed ledger, through a unique chain block structure to verify and store data, so as to ensure the privacy and security of all kinds of information, but the blockchain personal identity authentication link is not complete, relying only on the key can operate personal accounts there are serious security problems. Therefore, a signature scheme is proposed, which takes face biometric as input, uses convolutional neural network (CNN) facenet to encode the face feature information, and then uses homomorphic encryption scheme to encrypt the face encoding and compare it with the face template in the database. At the same time, in order to ensure the authenticity and security of opeation, the user’s rPPG signal are detected in real time during the authentication in considred successfully. Finally, using the information mixing algorithm, biometric and RSA key are fused to form a combined key for signature. The experiment shows that, under the condition of obtaining biometric information, the user’s identity is verified correctly, and the contract is signed correctly within 2 s. In the whole scheme, the template creation time is 5.62 s, the encryption time of the input biometric information is 0.52 s, the heart rate detection time (including the camera time) is 5.59 s, and the user can be fully identified within four times, with an accuracy of 98%. The scheme improves the security of the blockchain transaction and signing process.