【摘要】 描述了一种基于组合质量评估参数及支持向量机的音频隐藏信息盲检测方法。该方法对一系列音频质量评估参数进行组合,用多维特征选取的方法选择组合质量评估参数的种类和数量,生成支持向量机盲检测仿真模型,可对可疑的音频进行隐藏信息盲检测。该方法经计算机仿真,对DSSS和常用的变换域信息隐藏方法有较好的检测效果。更多还原

【Abstract】 An audio steganalysis was described based on integrated quality metrics and support vector machine (SVM). The blind detection simulating model with SVM was created with integrated audio quality metrics, which were chosen by multi- dimension feature extraction, and the probably stego-audio could be detected. Simulation results indicate that it generally does well in DSSS and transform domain.更多还原