Article Preview
TopIntroduction
The main feature of reversible data hiding is that the original cover media can be reversibly recovered after the hidden information is extracted. That is a great improvement in the field of data hiding and watermarking. Due to the reversibility, reversible data hiding can be used in a lager field such as medical image surgery, military imagery, and remote sensing imagery and so on. It is worthy to study reversible data hiding scheme, and many researches have been made in recent years.
There are three main categories for reversible data hiding schemes, one is compression based (Fridrich, 2001), the second one is difference expansion based (Tian, 2003; Tian, et al., 2004; Alattar, 2003, Alattar, 2004) and the third one is histogram modification based (Ni, et al., 2003; Ni, et al., 2006). Reversible data hiding based on compression makes use of the redundancy of cover image and thus the character of cover image limits the capacity and quality. Reversible data hiding scheme based on difference expansion was firstly proposed by Tian (2003), it hid one-bit information by extending the difference between two neighbor pixels. Alttar (2003, 2004) improved Tian’s scheme by extending n-1 pairs of neighbor pixels’ differences to hide n-1 bits information, and thus enhanced the hiding capacity. However, with larger hiding capacity, these kinds of schemes do not perform well in keeping the image quality. The hiding capacity increases while the quality of the image with hidden data drops quickly. Schemes based on histogram modification cause less distortion to the cover image and achieve better image quality. However, the obvious drawback of these histogram modification based data hiding schemes is that the hiding capacity is limited to the peak point of the histogram (Ni, 2003). In order to increase hiding capacity, there are two measures: raising the peak points’ height of the histogram or increasing the number of peak points. Quite a lot of schemes based on the two ways have been proposed: Lin, et al. (2008) proposed a multi-level embedding scheme to increase the capacity of histogram modification based scheme. Some prediction-difference schemes were also proposed to generate the histogram for data hiding. Tsai (2009) proposed a prediction model to get the prediction errors for histogram modification based scheme through explore the similarity between neighbor pixels. Kim (2009) sampled the original image to get a predicted image based on these sampled images; then calculated the difference images between the predicted image and these sampled images to get the histogram for data embedding. However, the hiding capacity of histogram based method is still not large enough compared with other schemes.