1D Frequency Transformation- Based Amplitude- Only Images for Copyright- and Privacy- Protection in Image Trading Systems

Main Article Content

Wannida Sae-Tang
Shenchuan Liu
Masaaki Fujiyoshi
Hitoshi Kiya

Abstract

This paper proposes an effective method of generating amplitude-only images (AOIs) which are inversely transformed amplitude components of images and which are used in the copyright- and privacyprotected digital fingerprinting-based image trading systems. The proposed method applies a one-dimensional frequency transformation to an image for generating the AOI with low intensity range (IR), whereas the IR of the AOI in the conventional method using a two-dimensional transformation is too large to store or transmit. The proposed method is distinct from other IR reduction approaches such as clipping, linear and non-linear scaling, block division, and random sign assignment which require extra-calculations, it reduces the computational complexity while it reduces IRs of AOIs. Moreover, experimental results show that the proposed method enhances the quality of fingerprinted images and fingerprinting performance.

Article Details

How to Cite
[1]
W. Sae-Tang, S. Liu, M. Fujiyoshi, and H. Kiya, “1D Frequency Transformation- Based Amplitude- Only Images for Copyright- and Privacy- Protection in Image Trading Systems”, ECTI-CIT Transactions, vol. 8, no. 2, pp. 133–144, Apr. 2016.
Section
Artificial Intelligence and Machine Learning (AI)