Research On Image Steganalysis Technique

With the rapid development of computer science and web multimedia technology, Information Security gets more and more attention. Steganalysis is a new branch of information security, which mainly researches on the effective detection of the cover image and boosts the development of information security. Especially after 9.11 event, steganalysis became more important, which plays an important role in national defence and military security. From the former LSB-based method to the universal algorithms, steganalysis has got great development in recent years, and provides maturely schemes for the information detection. Therefore, research on steganalysis is very valuable both in academic development and potential applications.In this dissertation, we firstly introduce the basic theory and development of Information Hiding and Information Security. The concept and the theory models of steganography and steganalysis are presented subsequently. Then we focus on the state-of-art of steganalysis and point out the current problems. Based on the above study, we mainly research on blind/universal steganalysis technique, and achieve some valuable results as follows:(1) By analyzing the changes of the DCT coefficients before and after stegnography, a steganalysis method based on the statistical characteristics of image DCT coefficients and SVM is proposed. We firstly introduce image prediction technique, and set up the Gaussian model for image DCT coefficients. Then, the parameter of the Gaussian model is estimated by using the maximum likelihood function method. Furthermore, the mean and variance of the parameter are evaluated as the statistical feature vector for the classification. In the end, we utilize SVM as the classifier and receive good performance.(2) We research on the concept and the relationship of image spatial moment and frequency moments. Using the merits of moments, we propose a universal steganalysis algorithm based on the scaling of wavelet sub-bands moments. The algorithm firstly calculates the histogram of the image wavelet coefficients, then evaluates the absolute moment and scales the feature vector. Finally, the scaled moment vector is regarded as the feature vector for the classification. (3) Design the classifier. By analyzing the principle and the performance of the classifier, and the effectiveness of the feature vector. SVM is utilized as the classifier. Compared with other classifiers, such as "Nearest Neighbor" and then analysis the detecting results. The performance is very well.