Research On Reversible Data Hiding Algorithms Based On Interpolation Images

With the development of information technique, more and more application fields are implementing information management. Digital information has become a widely procession content in the daily work. Computer network provides people with a platform of information transmission and information sharing. However, the network is an open platform and the security problem of data during transmission is still a huge problem.Information security technology includes cryptology, watermarking, data hiding technology and message authentication technology, etc. Data hiding technology is a technology which hides secret information into the cover media, then transfer to the receiver. It can ensure the security of secret information. Reversible data hiding as an important branch of data hiding, and it has received extensive attention of researchers in recent years. It can not only get the secret information in the information, also can restore the cover information exactly. So it is widely used in military, medical, cultural relic protection and law, etc.In the paper, we focus on the research of reversible data hiding based on the interpolation image and the lossless information hiding algorithm based on interpolation prediction algorithm. We propose a interpolation algorithm based on parabola and a data hiding algorithm based on interpolation image. According to interpolation pixel as the predict value, then calculate the error value with the corresponding pixel of image. Then the secret information is embedded into the errors. It implements the reversible data hiding. Some works have been completed as follows.1. An interpolation algorithm based on the parabola. The original image whose size is mxn. A column is inserted between every two columns and a lineis inserted between every two lines. The interpolation image’s size is2mx2n. The pixels whose row and co lumn coordinates are even number are the original pixel. The rest of the pixels are inter polating pixel. According to the three consecutive original pixels which is in same directi on to get a parabola. The interpolation pixel value is a weighted sum of the original pix els around the interpolation pixel with different proportion and the parabola value. The e xperimental results show that the interpolation algorithm has low time complexity and th e interpolation image has higher PSNR value.2. A reversible data hiding algorithm based on interpolation image.5x5matrix in the interpolation image is a unit. Three original pixels are used in the same direction to form a parabola firstly. The interpolation values are calculated by the parabola, and the mean of two adjacent original pixels in the same direction is calculated too. Then the difference values between the interpolation value and the mean value are calculated, and the secret bits number is calculated by the difference. The secret bits are embedded into the interpolation pixels by changing the value of interpolation pixels in accord with the opening direction of the parabola. Experimental results show that the proposed algorithm achieves large embedding capacity and higher PSNR value.3. A new data hiding algorithm based on prediction difference. Interpolation pixels which are obtained by the parabolic interpolation algorithm are as predictive value of pixels. The difference between interpolation pixels and the pixel which is in the corresponding position in the original image are calculated, and get the difference matrix. According to the secret bits to modify the difference, changing or staying the same to embed secret information.The above studies have achieved good comprehensive performance (embedding capaci ty/PSNR value).