Robust Lossless Data Hiding by Feature-Based Bit Embedding Algorithm

Abstract

Recently, data hiding, or information hiding, plays an important role in data assurance. Generally speaking, data hiding techniques can be classified into steganography and digital watermarking (Cox et al., 2008; Shih, 2008). The marked images generated by the steganographic methods (Gu & Gao, 2009; Liu & Shih, 2008; Qu et al., 2010; Wang et al., 2010; Zhou et al., 2010; Fan et al., 2011) were prone to catch damage (by manipulations) and resulted in a failure extraction of the message. However, based on the spatial domain, the steganographic methods often provide a large payload with a good perceived quality. Major applications of the techniques can be found in private data saving, image tagging and authentication, and covert communications. On the other hand, the robustness performance with a limited payload is a key feature of digital watermarking approaches (Lai et al., 2009; Al-Qaheri et al., 2010; Lin & Shiu, 2010; Yamamoto & Iwakiri, 2010; Yang et al., 2010; Martinez-Noriega et al., 2011). Most of the robust watermarking approaches which based on the transform domain such as discrete cosine transform (DCT), integer wavelet transform (IWT), and discrete Fourier transform (DFT) can be tolerant of common image processing operations. Their usages can be found in owner identification, proof of ownership, and copy control. Note that conventional data hiding techniques were irreversible, namely, the host media can not be recovered after data extraction. To preserve or protect the originality of the valuable (or priceless) host media, for example, military or medical images, and law enforcement, the reversible data hiding schemes, also known as lossless data hiding schemes were suggested to achieve the goal. For some applications, it requires to completely recover the host media if the marked images remain intact, and to extract the hidden message when the marked images were intentionally (or unintentionally) manipulated by the third parties. But, most of reversible data hiding schemes (Tian, 2003; Alattar, 2004; Hsio et al., 2009; Hu et al., 2009; Tai et al., 2009; Wu et al., 2009; Lee et al., 2010; Xiao & Shih, 2010; Yang & Tsai, 2010; Yang et al., 2010, 2011) were fragile in the sense that the hidden message can be unsuccessfully extract even if a slight alteration to the marked images, not to mention the recovery of the host media. Several authors (Zou et al., 2006; Ni et al., 2008; Zeng et al., 2010) therefore proposed robust reversible data hiding algorithms to overcome the issue.