Detection Of Hidden Information For Image In Spatial Domain

Since 1990's, with the development and popularity of computer techniques and network communication, more and more people choose network as a mean of transmitting information. As a result, the information security of network has been a focus of attention. Now, as an important branch of information security, steganography and steganalysis become a hot spot of information technique field.Steganography doesn't change the basic characters and functions of a cover. In this precondition, using data redundancy, steganography invisibly embeds secret message into a digital media, such as images, audio and video. Steganalysis is a kind of attacking behavior against steganography algorithm, with the purpose to reveal the existence of secret message in the digital media.Based on home and abroad research of steganalysis, using images as cover, this paper mainly discusses detection of spatial LSB (Least Significant Bit) matching algorithm.A special steganalysis algorithm for LSB matching is proposed in this paper. From the perspective of bit plane correlation, this algorithm extracts features, uses support vector machine as a classifier to detect LSB matching. Experimental results show that the proposed algorithm has better detection results when the embedding rate of stego images is higher than 40%, and good stability.The steganalysis algorithm based on gray level run length matrix, is a general detection method for LSB replacement and LSB matching steganography. It extracts features from the gray level run length matrix to detect the hidden information with two LSB algorithms. In this paper, some improvements and additions are done for the original algorithm. First, the kernel function of support vector machine is adjusted to improve the detection rate for LSB replacement steganography; Second, the length of secret message is estimated to overcome the shortage of original algorithm.