Enhancing Steganographic Algorithm Based On Image Pre-processing

With the development of Internet technology,the forms and quantities of network multimedia information such as pictures,audios,and videos have become richer,providing more and more abundant carrier libraries for various steganographic algorithms.As major scientific research institutions and universities attach importance to network security,steganography has received more and more attention and has been developed more fully.The adaptive steganographic algorithm is one of the safest steganographic algorithms today.The traditional steganographic algorithm does not pay attention to the image content,while the adaptive steganography algorithm focus the embedding in areas that are not easy to model,such as the image texture area and edge area,to improve the ability to resist steganalysis.The adaptive steganographic algorithm is divided into two parts: distortion cost calculation and embedding.With the introduction of the STC encoding method that can achieve the upper bound of theoretical embedding efficiency,the design of the steganographic algorithm is simplified to the distortion function design.However,a single distortion function cannot measure the safety of all pixels,so there will be low distortion cost abnormal points in flat areas and high distortion cost abnormal points in areas with complex textures.In order to solve the above problem,this paper studies the image steganography enhancement algorithm based on image pre-processing.Through the enhancement of the cover image,it solves the two types of abnormal point problems in spatial image steganography and JPEG image steganography,and designs a general image steganography enhancement method based on image pre-processing.At the same time,an enhanced framework based on adaptive steganography framework is proposed.By adding cover image pre-processing,embedding cost readjustment,and postprocessing of stego images,the existing adaptive steganography framework is improved and the problem that existing steganographic algorithm's safety only depends on the distortion function is solved.Finally,according to the blocking artifact of JPEG image and the relationship between blocking artifact and the security of steganographic algorithm,a JPEG steganography enhancement algorithm based on blocking artifact is proposed,and a multi embedding cost combination method is proposed to better measure the security of each pixel through two kinds of embedding costs.The main research contents and contributions of this paper are as follows:1)The two types of abnormal points in the spatial steganography are studied,and the existence of the two types of abnormal points is proved through experiments.Based on the characteristic that the noise in the flat area is sensitive to small-scale filters and the characteristics that filters of different scales can extract texture details of different sizes,a spatial steganography enhancement algorithm based on multi-scale filters is proposed.By preprocessing the cover image,the texture details in the image are extracted and enhanced,and the enhancement of the noise in the flat area is reduced by adjusting the weight.Finally,through experiments compared with the existing mainstream adaptive spatial steganographic algorithms,it is proved that this method can effectively solve the two types of abnormal point problems in spatial steganographic algorithms and improve the security of the existing spatial steganographic algorithms.2)The similarity between spatial steganography and JPEG steganography is studied,and experiment is carried out to prove that there are two types of abnormal points in JPEG steganography.At the same time,based on the different computational domains of mainstream JPEG steganography for embedding cost,a general JPEG image enhancement method is proposed.Based on the research of spatial steganography enhancement algorithm,a JPEG steganography enhancement algorithm based on multiscale filter is proposed to solve the two types of abnormal point problems in JPEG image steganography.Comparison of mainstream JPEG steganography algorithms proves that this method can improve the security of existing JPEG steganographic algorithms.Finally,an enhanced framework based on the adaptive steganography framework is proposed to improve the existing adaptive steganography algorithm framework and solve the problem that the security of the steganographic algorithm only depends on the distortion function.3)The problem of blocking artifact in JPEG image and the relationship between blocking artifact and the security of JPEG steganography are studied.The blocking artifact in JPEG images is reduced by using a deblocking filter.Then,the combination method of multi-embedding cost in JPEG steganography is studied,and the security of the steganographic algorithm is improved by combining the embedding cost of the two distortion functions,and the best combination algorithm is selected by experiments.Finally,based on the above research,a JPEG steganography enhancement algorithm based on blocking artifact is proposed,and compared with the existing mainstream JPEG steganographic algorithm through experiments.Experimental results show that this method can effectively improve the security of the existing JPEG image steganographic algorithm.