Enhanced high capacity image steganography using discrete wavelet transform and the Laplacian pyramid

Abstract

This paper introduces a high capacity image hiding scheme with enhanced stego image quality. This new hiding scheme utilizes a multiscale Laplacian pyramid of the cover image in the Discrete Wavelet Transform (DWT) domain. Previous work either enhanced capacity at the expense of stego quality or improved stego quality albeit at lower capacities. The proposed scheme will utilize the high-frequency bands of the DWT of the cover image for increased hiding capacity while further extending the payload capacity by hiding in the lowest level Laplacian pyramid of the DWT low-frequency band using a curve-fitting adaptive region approach in the spectral magnitude discrete cosine transform domain. The proposed scheme results in enhanced visual fidelity as well as high capacities as compared to competing methods. Comparative experimental results will show that the proposed scheme outperforms recent methods in terms of payload capacity as well as various image quality measures. Resistance to data-loss and noise tampering, geometric distortions and Checkmark attacks, and steganalysis detection attacks will further demonstrate the robustness of the proposed scheme.