Study On Adaptive Steganography Algorithm Of Digital Image In Spatial Domain

The development of Internet and information technology brings convenience to People's Daily life as well as various problems.At the same time,in the information age,information security is more important to countries,governments and individuals than ever before.In this case,people put forward higher requirements for the protection of important information,and steganography as an important research direction in the field of modern information security has been widely concerned by national and international researchers.At the same time,due to the wide spread and use of images on the Internet,image steganography has become the focus of research.In this dissertation,the adaptive steganography algorithm of digital image in spatial domain is studied.The main works includes two parts: the adaptive steganography algorithm based on clone selection and the adaptive steganography algorithm based on additive distortion function.1)The adaptive steganography algorithm based on clone selection establishes the immune information hiding model with the help of the concept of antibody in artificial immune system.The model,the image sharpening in image processing method is use to perform immune optimization on the cover image,so as to obtain more suitable cover image for steganography.in other words,embedding information in the new image information is much more difficult to be accurately detected by the detector than embedding information in the original cover image.To improve the security of steganographic algorithm.According to the image sharpening parameters,the appropriate antibody was designed for immune processing,and then the clonal selection algorithm was used to search out the optimal antibody for different cover images.This optimal antibody corresponds to the optimal parameter value of image sharpening.Then the best parameter value is used to sharpen the cover image to generate a new cover image,and the existing adaptive steganography algorithm is used to embed the information in the new cover image.Experiments show that the proposed method can improve the security of the existing algorithm without reducing the hiding capacity,and the method can be combined with a variety of adaptive steganography algorithms,which has strong practicability.2)The adaptive steganography algorithm based on additive distortion function designs a new additive distortion function based on the minimum embedded distortion model.When calculating the distortion cost of pixels,the distortion function comprehensively considers the correlation of the pixel neighborhood in the horizontal,vertical and diagonal directions.The stronger the correlation is,the higher the distortion cost will be.Firstly,the correlation degree of the pixel neighborhood is defined to quantify the correlation of different directions.For a pixel neighborhood,the value with the largest correlation in different directions is used to represent the correlation strength of the neighborhood,and the inverse of the maximum value is taken as the distortion cost of the pixel.At the same time,according to the principle of distortion diffusion,the calculated distortion cost is smoothed by means of an average filter.Finally,according to the distortion cost,the information is embedded with STC under the framework of minimizing embedding distortion.The experimental results show that the proposed distortion function can effectively avoid embedding secret information into the edge region in a single direction,and has high security in the face of two kinds of high-dimensional statistical feature detection,SRM and max SRM.