Skip to Main content Skip to Navigation
Theses

Content privacy and access control in image-sharing platforms

Abstract : In recent years, more and more users prefer to share their photos through image-sharing platforms. Most of platforms allow users to specify who can access to the images, it may result a feeling of safety and privacy. However, the privacy is not guaranteed, since at least the provider of platforms can clearly know the contents of any published images. According to some existing researches, encrypting images before publishing them, and only the authorised users who can decrypt the encrypted image. In this way, user’s privacy can be protected.There are three challenges when proposing an encryption algorithm for the images published on image-sharing platforms: the algorithm has to preserve image format (e.g. JPEG image) after encryption, the algorithm should be secure (i.e. the adversary cannot get any information of plaintext image from the encrypted image), and the algorithm has to be compatible with basic image processing in each platform. In this thesis, our main goal is to propose an encryption algorithm to protect JPEG image privacy on different image-sharing platforms and overcome the three challenges. We first propose an encryption algorithm which can meet the requirements of the first two points. We then implement this algorithm on several widely-used image-sharing platforms. However, the results show that it cannot recover the plaintext image with a high quality after downloading the image from Facebook, Instagram, Weibo and Wechat. Therefore, we add the correcting mechanism to improve this algorithm, which reduces the losses of image information during uploading the encrypted image on each platform and reconstruct the downloaded images with a high quality.
Complete list of metadatas

Cited literature [110 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-02873678
Contributor : Abes Star :  Contact
Submitted on : Thursday, June 18, 2020 - 2:33:06 PM
Last modification on : Monday, October 5, 2020 - 9:44:04 AM

File

2017_HE_archivage.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-02873678, version 1

Citation

Kun He. Content privacy and access control in image-sharing platforms. Social and Information Networks [cs.SI]. CentraleSupélec, 2017. English. ⟨NNT : 2017CSUP0007⟩. ⟨tel-02873678⟩

Share

Metrics

Record views

102

Files downloads

46