Research On Information Hiding Algorithm Based On Digital Image

With the development of computer technology, the communications in network increase largely, information security has become a hot topic of concern to people. In the new situation, the encryption technology alone has been unable to meet the needs of information security. Information hiding technology is new areas of information security, with some great advantages which the traditional encryption technology can not match, in the hidden aspects of communication and computer network forensics information hiding are playing an increasingly important role, and has broad prospects for development.More and more problems appear in information security today, which means encryption technology alone can not fully solve the problems in the information security. So people use the information technology in the information security area, providing a more secure and reliable means of communication. Information hiding, also known as data hiding, use the characteristic of human sensory, hide the secret information into the carrier part of the feeling of redundancy, other people can not found the existence of hidden information only use eyes.This paper reviews the theory and knowledge of information security and information hiding, and introduces the basic concepts and principles of digital image information hiding technology. Then for a carrier of digital image information hiding technology, proposed the following three types of hidden algorithm:1. Proposed a data hiding algorithm based on matching:using the method of bits matching to find the most similar part of the cover image for embedding the secret information, so that the amount of changes become smallest. First, proposed the whole matching algorithm and the overall group matching algorithm for embedding information into a single bit plane, also using the threshold filter and the second group of matching algorithm to resolve the problem of slow running time. Then analysis the situation changes when the embedded information into multi-bit plane, modify the three aspects of extraction methods in the carrier string, matching the target, cover image adjustment, proposed a multi-bit plane matching algorithm based on the original algorithm. Experimental results show that the proposed algorithm can effectively reduce the distortion of the cover image while large amounts of information embedded, also to maintain a high PSNR. 2. Proposed a information hiding algorithm base on wavelet variability and resisting RS analysis:Algorithm combine the block classification algorithm based on wavelet variability and the dynamic compensation that can resist RS analysis, first divide the image into 8×8 sub-blocks, using the wavelet energy calculation sub-block variability, and use the masking properties of human eyes to adjust the depth of the embedded block according to the variability of each block to enhance image quality; in the other way with flipping each row to reduce the RS value so can reach the purposes of resisting RS analysis. Using PSNR and RS value to construct the fitness function, and then calculate adaptively the optimal embedded parameters that can simultaneously achieve better visual effects and the performance of resisting RS analysis. The experimental results show that the algorithm can maintain a high PSNR and significantly reduce the RS value for many cover images, while embedding large-capacity information.3. Proposed an information hiding algorithm that can resist salt and pepper noise:When the host image destroyed by salt and pepper noise, the extracted hidden image contains a more complex form of noise which is not easy to accurately detect. Hide images directly to repair damaged, difficult to achieve satisfactory results. Therefore, detect the relatively simple salt and pepper noise from cover image, then locate the hidden bit may be damaged, and use the gray value of the neighborhood points to repair these bits. Firstly, since the gray values of the salt and pepper noise point is close to 0 or 255, we can use this property to detect the noise pointssimply. Extract information bit and classify into destroy bit or correct bit by is extracted from the noise point or not. Then use of information bits to pixels of the image reconstruction. Estimate the value of the center point with gray values of neighborhood points, and use the estimated values to correct the destroyed bits. Experiments show that this algorithm can be salt and pepper noise in the image carrier after the destruction of a relatively complete extract and repair the hidden image, and compare to other filtering algorithms the restored hiding image has better visual quality compared.