Research On Information Hiding And Hiding Image Restoration Algorithm

With the rapid development of computer science and technology, more and more information can be transferred by means of network. It brings about great convenience for people, but imperil the safety of information at the same time. The secret information is hided into public carrier by information hiding that can realize safe transmission of information. Whereas, stego-image will be attacked by noise, cropping when stego-image is transferred on network, to some extent, it also threaten the safety of information.In recent years, information hiding has been improved greatly with the development of chaos technology. Many people apply chaos theory in information hiding algorithm in order to enhance safety of algorithm. Until now, there is no comparatively perfect resisting attack1information hiding algorithm in domestic and foreign research, so information hiding for resisting attack still has very large development space. There is still very large development space for information hiding for resisting attack.Our research is information hiding algorithm based on image. The main research content of our paper is to improve the ability of resisting cropping and noise and achieve reversible information hiding. Main work is shown as follow.1. We propose a watermarking algorithm for resisting cropping attack based on Haar Wavelet Transform. The original watermarking is divided into four equal pieces after it is scrambled and encrypted. Then we get three new watermarking images which are generated by the four equal pieces. The original watermarking and three new watermarking images are embedded into the four components of2-level wavelet transform (WT) for the cover image respectivly. For the low frequency component, we convert tens digit of low frequency coefficients to a binary. The secret is embedded into the low frequency component by replacing the lowest bit of the binary. For the middle and high frequency component, we embed secrets according to the size relationship of the two adjacent coefficients. Experimental results show that the proposed algorithm is robust against noise attacks and large area cropping attacks.2. We propose lossless high capacity data hiding algorithm based on neighboring pixels. We transfer cover image into one dimension sequence according to three different scanning methods. We can get three difference sequences by calculate two adjacent pixels of the three sequences. Total times of two values which appear most in three difference sequences respectively will be obtained. We choose the scanning method by which maximal time can be obtained. Last we can get two values which appear most according to the scanning method we choose. The secret can be hided between the two values. We can determine how many bits can be hided into every value according to our need. Experimental results show that the propose algorithm achieves a high embedded capacity and still maintains a high stego-image quality.3. We propose an image restoration algorithm for information hiding based on bit restoration. We mark every bit of pixels of stego-image for credible bit and incredible bit. First, we restore the incredible highest bit by applying credible high bits of neighbor pixels. Then two shortest credible bits which are equal are used to restore incredible bit that is between the two shortest credible bits. Last, we use valuation method of restoration to restore every pixel. Experimental results show that our algorithm can recover secret image well even if the stego-image is in high noise.