Association Analysis and Content-Based Information Hiding Approach with High Hiding Capacity

In this paper, we present a novel content-based information hiding scheme which analyse the association between cover image and secret image. In order to increase the hiding capacity and imperceptibility, we plan to embed a less energy image called residual image instead of the original image by a spatial domain method least signification bit (LSB), which obtained from the statistical method-partial least squares (PLS) and genetic algorithms (GA). Furthermore, Arnold iteration transform is applied to increase the security when secret image is hiding into the cover image. As our experimental results, the capacity and performance of the proposed method is better compared with the original method which lose sight of association between cover image and secret image.