Reversible data hiding in enhanced images with anti-detection capability

Reversible data hiding (RDH) technique has been widely used for content authentication. However, most conventional RDH schemes generally concentrates on the rate-and-distortion performance (namely high capacity and low distortion), instead of their undetectability performance (namely security). In other words, the stego images with conventional RDH schemes could be easily detected by most steganalysis tools, e.g., SPAM, to reveal some suspicious secret data hidden in it, which increases the embedding risk. Therefore, their applications in some circumstances are restricted. In this paper, enhanced images generated by common image enhancement techniques are built as a desired carrier for RDH, to achieve both better rate-and-distortion and desired undetectability performance due to its special statistical histogram distribution, where many vacant interspaces are scattered in the histogram and suitable for histogram shifting based RDH process. In addition, considering that the enhanced images could provide much more candidate spaces for embedding, some strategies for security improvement, such as multi-feature sorting technique and so on, could be introduced. Experimental results verify the superiority of proposed scheme compared with other related algorithms.