Cryptology ePrint Archive: Report 2018/1202

AuthCropper: Authenticated Image Cropper for Privacy Preserving Surveillance Systems

Jihye Kim and Jiwon Lee and Hankyung Ko and Donghwan Oh and Semin Han and Kwonho Jeong and Hyunok Oh

Abstract: As surveillance systems are popular, the privacy of the recorded video becomes more important. On the other hand, the authenticity of video images should be guaranteed when used as evidence in court. It is challenging to satisfy both (personal) privacy and authenticity of a video simultaneously, since the privacy requires modifications (e.g., partial deletions) of an original video image while the authenticity does not allow any modifications of the original image. This paper proposes a novel method to convert an encryption scheme to support partial decryption with a constant number of keys and construct a privacy-aware authentication scheme by combining with a signature scheme. The security of our proposed scheme is implied by the security of the underlying encryption and signature schemes. Experimental results show that the proposed scheme can handle the UHD video stream with more than 17 fps on a real embedded system, which validates the practicality of the proposed scheme.

Category / Keywords: cryptographic protocols / Privacy, Authentication, Forward-secure signature, Video

Date: received 13 Dec 2018

Contact author: easyone518 at gmail com

Available format(s): PDF | BibTeX Citation

Version: 20181218:193910 (All versions of this report)

Short URL: ia.cr/2018/1202


[ Cryptology ePrint archive ]