Ciphertext jpeg Image Retrieval Method Based on Tree Bow Model

Abstract

The invention discloses a cipher text JPEG image retrieval method based on a tree-form BoW model. The method comprises the following steps: firstly, an image owner encrypts all images in a database byadopting a double-layer encryption algorithm and uploads a cipher text image database to a cloud server, wherein the encryption means is composed of DCT coefficient stream encryption and scrambling encryption; secondly, the cloud server calculates the DCT coefficient statistical characteristics of cipher text images, adopts the statistical characteristics as local characteristics, and convertingall local characteristics of the images into global characteristics according to a tree-form BoW model; finally, the server inquires the images according to a cipher text submitted by an authorized user, extracts local characteristics, converts the local characteristics into global characteristics, calculates the distance between the global characteristics of each image and the global characteristics of a database image, then judges the similarity, and returns an obtained result to the user. Based on the method, during the image retrieval process, the privacy of images and characteristics canbe protected at the same time. The characteristics extraction is completed by the server. According to the invention, the calculation burden of a user is reduced. The transformation of the tree-form BOW model on characteristics ensures the high retrieval efficiency.

Description

technical field [0001] The invention relates to a ciphertext image retrieval method, in particular to a secure ciphertext image retrieval method in a cloud environment, which is applied in the technical fields of information retrieval methods and image encryption protection. Background technique [0002] Cloud computing is the most popular service model nowadays. Users can obtain the required resources through the network anytime and anywhere without considering other factors such as hardware. Therefore, more and more enterprises and individuals choose cloud servers to store and manage local data. However, this also brings hidden dangers to the privacy and security of data. One solution is to encrypt the data and then upload it to the cloud server, which is a huge challenge to traditional plaintext-based information retrieval services. How to achieve accurate retrieval while protecting data privacy, such as retrieval technology based on ciphertext images, has become a res...

AI Technical Summary

Problems solved by technology

[0006] In the first type of method, the extraction of image features needs to be completed by the owner and the user, and the image and features need to be encrypted and stored separately, which increases the computing burden of the user and the storage overhead of the server

In the second type of method, on the one hand, the unencrypted features leak the private information of the image, and the server is likely to infer the plaintext content of the image based on the plaintext features, thus threatening the security of the image content

On the other hand, the retrieval method of judging image similarity by comparing local features one by one is relatively inefficient.

Images