Description

This paper presents a highly secure data hiding system (a.k.a. Steganography) in a cover image using the low frequency Curvelet domain. The contribution of the suggested technique is its high security, because it is using four secret keys (encryption key, two shuffling keys, data hiding key) and using only the low frequency component of Curvelet domain. The use of low frequency component of Curvelet transform in steganography provides a number of advantages compared to other techniques such as: 1) Computation time reduction and, 2) Curvelet transform are designed to handle curves discontinuities using only a small number of coefficients, so hiding in the low frequency components will not affect edges coefficients, which produces better stego object quality. Robust Image Adaptive Steganography using Integer Wavelets was introduced. The algorithm offers excellent performance without any error correction codes. It can hide high volumes of information compared to many existing robust and secure data hiding techniques and yet retain better values of peak signal to noise ratio (PSNR) and MSE. The algorithm requires hiding of the position memory along with the payload in the stego image which may be a drawback in certain cases, but still have some problems in the performance of the stego image like handling the discontinuity and sudden changes in image edges.