Research On Adaptive Steganographic Distortion Function For Binary Image

With the increasing development of information technology and the popularization of the Internet,the information transmission in daily life is getting more convenient and frequent;meanwhile,it brings a lot of risks.Information storage or transmittion in the network is susceptible to be stolen by attackers.In order to protect the security of these data,encryption technology is usually used to produce unrecognized ciphertext by encoding the secret message.Although this method can protect the data information,it is prone to attract the attention of attackers due to the messy and meaningless characteristic of the ciphertext.Steganography can solve this problem well that it not only helps to hide the specific content of secret messages,but also protects the covert communication behavior.There are many carrier types in steganography,such as image,audio,video and so on.Binary image is a common and widely used image type,including fax file,handwritten image,medical image,halftone image and so on.The halftone image is a special type that the original continuous image is processed by halftone technology.At present,the mainstream steganography algorithms are based on the framework of minimal-distortion model,where the distortion refers to the security cost of modifying the carrier elements,so the performance of a steganography algorithm largely depends on the design of distortion function.Because the binary image has just black and white pixels,there is less redundant space in the image and it is very sensitive to the modification of pixels,bringing about great challenges to steganography.To solve this problem,this paper focuses on how to design a binary image adaptive distortion function with better performance.The main work content and innovative points of this dissertation are as follows:1.Design method of Non-additive Distortion Function for Binary Image based on Layered EmbeddingIn this paper,a binary image steganography method with non additive distortion is proposed.The method divides the carrier image into two sub images,updates the distortion by transferring the influence factor,and realizes the message embedding by using the minimum distortion steganography coding.Compared with the previous aditive distortion steganography method with fixed distortion,this method combines the characteristics of binary image with only blac k and white pixels,so that it can capture the modified correlation between pixels and realize dynamic modification distortion,so as to improve the visual quality and security of binary image steganography.2.Design method of Halftone Image Distortion Function based on Side InformationThis paper proposes a distortion definition method for halftone image that compares the original continuous image with the binary images processed by halftone technology,extracts the side information which is beneficial to steganography,generates the influence factor,and guides the embedding of secret message through combining with the original distortion.Compared with the previous steganography algorithm without side information,the steganographic image generated by this method can retain the information of the original gray image to a greater extent,leading to improve the visual quality and security of halftone image steganography.