Image Steganalysis Algorithm Based on Attribute Dependent Degree
YU Wen-qiong · Journal of Sanming University · 2010
The accuracy of hidden information blind detection in which rough set attribute reduction is adopted would drop to some extent in ref[6].Steganalysis algorithm based on decision attribute dependent degree is proposed.In this algorithm,improved measures for the discretization of the decision table are proposed to get the best decision ability of discretized decision table with the detection accuracy improved,according to the knowledge of decision attributes dependent degree in rough set theory.Experiments are taken to Cox,Piva as the following order: firstly,optimize the decision table by this algorithm;secondly,get the minimal feature after attributes reduction;thirdly,build the classifier by SVM.The experimental results shows that the detection accuracy and efficiency have been greatly improved by the help of the proposed algorithm.