Privacy-protecting biological characteristic image processing method and device

Abstract

The embodiment of the invention provides a biometric image processing method and device, and the method comprises the steps: firstly obtaining a to-be-coded first image, segmenting the first image into a first number of first image blocks, and sequentially numbering the first image blocks; according to a number sequence in a preset first sequence, and disrupting the arrangement of the plurality offirst image blocks to obtain a second image which is arranged out of order; enabling the neural network model running on the server to perform biological feature recognition on the second image, wherein the neural network model is obtained by training specified training samples in advance, and each training sample in the specified training samples is arranged out of order according to a first sequence. The method can effectively improve the security of the private data of the user.

Description

technical field [0001] The embodiments of this specification relate to the technical field of privacy data protection, and in particular, to a privacy-protecting biometric image processing method and device. Background technique [0002] Biometric Identification Technology uses the inherent physiological or behavioral characteristics of the human body for personal identification. Biometric features are not easy to forge and lose. Compared with traditional authentication methods such as passwords and USB shields, identity authentication based on biometric features is more convenient and has a very broad application prospect. Among them, biometrics, as a kind of private data that can be used to identify user identities, require a higher level of security protection to prevent leakage. [0003] At present, in many identity authentication schemes based on biometric features, neural network models are used to identify biometric features. These schemes generally encrypt and prot...

AI Technical Summary

Problems solved by technology

These schemes generally encrypt and protect the collected biometric information during network transmission and storage, but in the neural network model identification link, it is necessary to input the original biometric information in an unencrypted state to the neural network. At this time, the original biometric information is The difficulty of stealing is relatively low, there is a risk of leakage, and the security of user privacy data needs to be further improved

Images