Lossless Data Hiding Using Optimized Interpolation Error Expansion

Abstract

Reversible data hiding technique enables the exact recovery of original contents upon extraction of embedding data. Many conventional watermarking algorithms that have been proposed so far, suffer from the problem of low embedding capacity. In many applications huge amount of data is required to embed and exact recovery of host image is also needed. Luo's additive interpolation-error expansion technique [1] is a high capacity reversible method for data embedding without zero loss of host information. In this paper, we focus on improving payload and imperceptibility, and we propose an algorithm which is an extension of interpolation-error scheme using Particle Swarm Optimization (PSO) which better exploits estimation of missing pixel. The experimental results demonstrate that due to slight modification in interpolation-error technique, we achieve large hiding capacity and improved image quality compared with other schemes.