Research And Implementation On Image Adaptive Steganography And Steganalysis

Digital steganography is an art of covert communication which embeds secret messages into seemingly normal carrier file.Now,Digital image is one of the most widely used carrier file types in digital steganography and many digital image steganography algorithms have been proposed.However,the definition of image embedded distortion,steganographic algorithm design and so on still have shortcomings.In addition,most work focused on the design of steganography algorithm,the corresponding software development cannot be discussed.In this paper,the image steganography and steganalysis detection technology were studied and implemented,the main research works and results are as follows:1.Aiming at some problems in the research of digital image steganography,this paper makes a deep research on the adaptive image steganography technology,and proposes a new image content adaptive steganography algorithm,and obtains better anti-detection ability.For the design of new adaptive steganography algorithm,the image is decomposed in multiple scales and directions by the Gabor filters,and then the influence of the image pixel or DCT coefficient on the image filtering coefficient is analyzed.Finally,the double-layer STC(Syndrome-Trellis Codes)coding is used to minimize the embedding distortion function value while embedding and extracting the secret message.Since the Gabor filter can capture the changes caused by steganographic embedding to image texture features from multiple scales and directions,the proposed steganographic algorithm can adaptively constrain the embedding changes to complex texture areas that are difficult to model.2.The image steganography software system is developed based on the proposed image steganography algorithm.For the development of steganography software,the different problems were considered to make the steganography safer.For example,the steganography technology is combined with encryption method,the multiple security mechanisms were utilized.In addition,the secret messages can be embedded continually in multiple file and it can better meet the needs of the actual image steganography..3.The steganalysis method for adaptive JPEG steganography was studied,one method for steganalysis feature extraction is proposed based on the entropy values of 2 dimensional(2D)Gabor wavelets.The 2D Gabor wavelets have certain optimal joint localization properties in the spatial domain and in the spatial frequency domain.It can describe the image texture features from different scales and orientations,at the same time,the entropy values of image filtering coefficients are usually used to characterize the texture of image.Therefore,the combination of the 2D Gabor wavelets and entropy can capture the changes of image texture characteristics caused by steganography embedding effectively.As to the proposed feature extraction method,the decompressed JPEG image is filtered by 2D Gabor filters with different scales and orientations firstly.Then,the entropy features are extracted from all the filtered images.Lastly,the ensemble classifier is used to assemble the proposed steganalysis feature as well as the final steganalyzer.The experimental results show that the proposed steganalysis feature can achieve a competitive performance by comparing with the state of art steganalysis features when they are used for the detection performance of the latest adaptive JPEG steganography such as UED,JUNIWARD and SI-UNIWARD.Last,the work of the whole paper is summarized and the future researches are also discussed.