摘要
常见的采用高斯核支持向量机(Gaussian support vector machine,G-SVM)分类算法构建分类器的隐写检测方法对最低比特位(Least significant bit,LSB)匹配隐写算法均存在训练时间过长的问题。针对这一问题,提出一种改进逻辑回归分类算法,即L曲线截断正则化迭代重加权最小二乘(L-curve truncated-regularized iteratively re-weighted least squares,LTR-IRLS)算法。该算法采用L曲线法来确定适合于隐写特征的Tikhonov正则算法的近似最优参数,并通过实验寻找出符合隐写特征的截断牛顿算法收敛参数,从而提高了检测准确率;采用重加权最小二乘法计算最大似然估计,并通过截断牛顿法避免计算最小二乘中的海森矩阵,降低了计算量。理论分析与实验结果证明,针对LSB匹配隐写检测,LTR-IRLS分类算法在保证检测准确率优于G-SVM分类算法的情况下,极大地降低了训练时间,从而提高了检测速度。
Least significant bit（LSB）matching algorithm and common steganographic methods,which use Gaussian support vector machine（G-SVM）algorithm as the classifier,spend too much training time.Therefore,an improved logistic regression classifying algorithm named L-curve truncated-regularized iteratively re-weighted least squares（LTR-IRLS）is proposed.Firstly,near-optimal parameters of Tikhonov regularization are determined based on L-curve,and convergence parameters of the truncated Newton algorithm are obtained through experiments for increasing the detection accuracy.Secondly,iteratively reweighted least squares are utilized to search for the maximum loss expectancy and truncated Newton methods are utilized to avoid computing the Hessian matrix in the objective function,therefore reducing the computation amount greatly.Theoretical analysis and experimental results verify that LTR-IRLS can ensure the detection accuracy rate higher than G-SVM classifier,meanwhile reducing the training time and increaseing the detection speed.
出处
《数据采集与处理》
CSCD
北大核心
2015年第6期1160-1168,共9页
Journal of Data Acquisition and Processing
基金
天津市自然科学基金(15JCYBJC15500)资助项目
关键词
L曲线法
迭代重加权最小二乘
截断牛顿法
隐写检测
LSB匹配
L-curve
iteratively re-weighted least squares（IRLS）
truncated Newton methods
steganalysis
least significant bit（LSB）matching
