Research On Steganalysis Of Images And Videos Based On Statistical Features

Starting from the 90s of last century, information hiding technology has become an important academic branch of information security. Battles between steganography and steganalysis have become an important issue in information hiding, which makes steganography and steganalysis restrict mutually, and promote mutually. This thesis is focused on the techniques of steganalysis in images and videos, and the practical applications of steganalysis techniques. Two steganalysis algorithms against image steganography and one algorithm against video steganography are presented. The main content of this thesis is summarized as follows:1. In this paper we analysis the statistical characteristics of original images and video frames. The focus of my study is to explor the statistical characteristics of neighboring pixels and the distribution of 01 pixels of bit plane. These results can provide the basis for the steganalysis.2. Bit-plane complexity segmentation (BPCS) steganography replaces bit-plane pixel blocks of high complexity with secret data. This steganography provides large embedding capacity and good imperceptibility. However, a security flaw in the BPCS method is revealedl. The continuity of the complexity histogram of the pixel block is affected by the embedded data. Based on the analysis of the security flaw in the BPCS steganography, a detection algorithm against BPCS embedding is proposed. The algorithm can reliably detect the existence of hidden data. It is easy to implement and computationally efficient, and is suitable for real-time detection. The experiment shows that our new arithmetic is more reliable than the existent one.3. A Block Complexity based Data Embedding (ABCDE) steganography replaces bit-plane pixel blocks of high complexity with secret data using new complexity measure. The principle of the method is the same as that of BPCS-Steganography. Embedding is performed by replacing pixel data of noisy regions in an image with another noisy data obtained by converting data to be embedded. The new complexity factors are more suitable to tell the difference between noisy regions from normal ones. But a security flaw is revealed by constructing a histogram of bit-plane block complexity. Based on the analysis of the security flaw in the ABCDE steganography, a detection algorithm against ABCDE embedding is proposed by analysis the discontinuous coefficient. The algorithm can reliably detect the existence of hidden data. It is easy to implement and computationally efficient, and is suitable for real-time detection.4. MSU(Moscow State University) Stego Video is a public video steganographic tool. It uses a specific data distribution mode to load secret data, so that the message can be extracted correctly even the stego video are compressed. In this paper, based on the analysis of the distribution characteristics of the embedded data, a detection algorithm against MSU embedding is proposed by utilizing the difference of block artifacts distribution between natural frames and stego frames. The algorithm can reliably detect the existence of hidden data. It is easy to implement and computationally efficient, and is suitable for real-time detection.