Secure image steganography using framelet transform and bidiagonal SVD

Abstract

Steganography and steganalysis are the prominent research fields in information hiding paradigm. This work presents a novel framelet transform based image steganography scheme that hides a secret image into cover image. Perfect reconstruction, sparsity, and stability enables framelet transform to be considered as suitable decomposition technique to obtain transform coefficients. The scheme also benefits from bidiagonal singular value decomposition. Secret information is embedded in singular values of framelet coefficients and stego is obtained. A variety of experiments is conducted to judge the efficacy of proposed method. Simulation results prove that stego images possess better visual quality and are robust to several popular image processing operations. Security performance of proposed method is investigated using various steganalysis schemes that include Gabor filter based, wavelet based and contourlet based steganalysis. Detection accuracy is found to be poor in all cases and confirms the undetectability.