The application of support vector machines in detection of images steganography
Xiaoyuan Yang, Zhigang Wang · Journal of Xidian University · 2005
The Support Vector Machines is a new machines learning algorithm based on the Statistical Learning Theory. It has found important applications in pattern discrimination of high dimensions Character vectors. The algorithm of Support Vector Macnines has been studied in this paper, and the detection algorithm of steganography based on the Support Vector Machines has been put forward. For this detection algorithm, we select two kinds of steganography software F5r11 and Jsteg4.1 to make a large number of experiments. In quadratic programming, we find that the penalty gene C is important to experimental results, and therefore, we show different detection results of different C's, and the results show that the detection rate of this algorithm improves evidently compared with the Fisher Linear Discrimination.