Intelligent cloud platform face recognition and local encryption method based on a neural network

Abstract

The invention discloses an intelligent cloud platform face recognition and local encryption method based on a neural network. The method comprises the steps of processing an original image through animage pyramid; establishing a face / non-face classifier, and outputting a large number of results which may be face classification, boundary frame regression and face position; carrying out bounding box regression, refining the candidate windows, and discarding a large number of overlapped windows; carrying out landmark positioning, removing a candidate window and displaying five face key point positioning at the same time; and encrypting the data through a chaotic logic diagram and an RC4 stream password. A local encryption mode is adopted, so that the picture encryption time is shortened, andthe encryption efficiency is greatly improved.

Description

technical field [0001] The invention relates to a face recognition and partial encryption method, in particular to a neural network-based intelligent cloud platform face recognition and partial encryption method. Background technique [0002] With the rapid development of the Internet and cloud storage technology, the wide application of big data and the popularization of smart terminals, the concept of synchronization and backup has gradually penetrated into people's work and life. Users can store and To manage data, in order to prevent data loss on terminal devices, users hope to back up data to a safe place, making it more convenient for users to access their own data on different terminal devices. The intelligent cloud service allows the mobile phone and the cloud to synchronize the cloud data of contacts, short messages, photos and other data. However, with the increase of personal information data, the leakage of cloud data due to hacker attacks and other reasons will...

AI Technical Summary

Problems solved by technology

However, with the increase of personal information data, cloud data is leaked due to hacker attacks and other reasons, which will bring serious personal privacy and other security risks to people. Personal information needs to be encrypted and stored on the intelligent cloud platform.

Considering that most of the pictures on the mobile phone are stored in high-definition, fully encrypting each high-definition picture will greatly increase the operating pressure on the cloud. Partial encryption can greatly reduce the length of encryption and the operating pressure of the cloud system. The template matching method usually used in traditional face recognition technology, the flexibility of the template itself is relatively poor, and it cannot be changed according to the change of the face. In real life, the face will be blocked due to lighting, wearing glasses, hats, postures, etc., or exaggerated expressions, which will directly affect the final detection. Therefore, it is necessary to design a face detection technology based on neural networks. Perform face detection, and then only encrypt the detected face area, and finally store it in the cloud, effectively solving the problem of personal privacy and security of photos in the cloud

Images