Secure and lossless data hiding by histogram shifting and scan paths using integer wavelet transform

摘要

This paper proposes a lossless image hiding method in the integer wavelet transform domain using histogram shifting in detail bands and scan paths in approximation band. This lossless algorithm can extract the hidden secret image without any distortion after the hidden data are extracted. Histogram shifting method is used to hide data in wavelet coefficients of high frequency sub-bands. It shifts a part of the histogram of high frequency wavelet sub-bands and thus embeds data by using the created histogram zero point. Since approximation part is the sensitive part of the cover image, we use difference histogram shifting method based on defined nine basic scan paths in low frequency sub-bands. This technique exploits the fact that gray scale values of adjacent pixels are close to each other. Also multilayer embedding is used to increase the hiding capacity.