Reversible Data Hiding Using Histogram Shifting for Digital Images

Abstract

Reversible Data Hiding (RDH) includes embedding information into cover media in a way that takes into consideration the recovery of the inserted information and in addition the original media. In our daily lives, JPEG is the most popular digital image format. Reversible data concealing technology has been frequently applied to JPEG images in recent years for specialized uses including management of file and for authentication of image. Moreover, because JPEG images have lesser information redundancy than uncompressed images, RDH in JPEG is far more difficult than in uncompressed images and any modifications made to the compressed domain could cause the host image to become even more distorted. In this study, we give some fundamental ideas to choose quantized DCT coefficients for RDH taking into account the statistical properties of DCT coefficients and theory behind the JPEG encoder. One of the most widely used methods for developing reversible data concealment technologies is histogram shifting. A new approach for JPEG images, histogram shifting-based RDH technique is put forth, zero coefficients are not extended; only coefficient values of +1 and -1 are used to convey message bits. In this research, we generate DCT coefficients for data hiding in an adaptive way by using a block selection strategy based on the number of zero coefficients in each 8x8 block. According to the experimental results, the recommended approach achieves a higher peak signal-to-noise ratio and has a smaller file size increase when compared to state-of-the-art approaches.