Research Of Image Steganography Based On Related Pixels

Steganographic technique is a new technology in the field of information security, it plays a very important role in privacy information, military affairs and the security of a country. The basic requirements for steganography algorithm are security and secret information embedding capacity. As the current steganalysis methods use the statistical analysis to detect the statistical abnormality of steganography digital media,the statistical imperceptibility is the main factor that steganography algorithm should take into consideration.The least significant bit based steganographic approach is easy to be implemented and has high embedding capacity,while the human visual system based steganographic algorithm has better visual imperceptibility and provides high embedding capacity. The two types of steganographic schemes therefore have been widely used. However, many steganalysis algorithms have been proposed to be against them. This paper first discusses the basic purposes of steganography and steganalysis,and their principles and main techniques, followed by the review of the current research status of the area. This paper also analyzes the previous steganographic method, and then proposes steganographic algorithms that are able to effectively resist statistical analysis and are of high security based on the statistical abnormality of the existing method. The main contributions of this paper are as follows:(1) Pixel feature steganography algorithm uses pseudo-random function MD5to generate pseudo random sequence, and to determine the secret information embedded position in the original image sequence. Then according to human visual characteristic, it divides an image into smooth area, transitional area, and marginal area by use of target pixels of the pixel difference value between neighboring pixels. It uses different bits substitution in three areas above.In order to improve the quality of the stego-image, we utilize the method of the optimal pixel adjustment process(OPAP). Experimental results indicate that the algorithm proposed can increase embedding capacity of image secret information, improve ability to resist statistical analysis, advance the accuracy of extracting secret information, and maintain a higher peak signal to noise ratio of stego-image.(2) Side matches(SM) based steganography cannot extract the embedded information exactly and furthermore has the shortcoming of statistical abnormality of the pixel difference histogram. Based on this,we proposed based on a modular function and edge matching steganography algorithm, algorithm firstly SM is used to determine the small pieces of image region, and then in different areas use different analog value of modulus function to embed secret information.Experimental results indicate that the algorithm proposed can increase embedding capacity of image secret information, improve ability to resist statistical analysis, advance the accuracy of extracting secret information, and maintain a higher peak signal to noise ratio of stego-image.To sum up, this paper has been completed a large number of experiments,implement and test the embedding and extraction process of the algorithm.. Experimental results show that our method provides a relatively good quality stego-image and the embedded secret-image can be completely recovered.