基于局部二值模式的中值滤波检测算法

摘要

针对数字图像处理及窜改中一种常用的处理模式———中值滤波，提出一种基于局部二值模式（LBP）的中值滤波窜改检测方法。首先对待测图像利用基于中心对称像素的局部二值模式算子（C-LBP）定位其特征提取区域，随后基于局部区域中像素之间的关系提出一种改进型的局部二值模式算子，并将其直方图统计作为最终的检测特征。实验证明，与其他传统滤波检测算法相比，该算子不仅大大降低了计算时间，而且在低维度的条件下取得了较高的检测率，是一种有效的中值滤波检测算子。%This paper proposed a novel algorithm for median filtering tamper detection in digital images which based on the lo-cal binary pattern (LBP)features.The algorithm firstly used center-symmetric local binary pattern (C-LBP)to locate detec-tion area in the suspected images,and then defined an enhanced LBP operator according to the relationship between adjacent pixels in the local.In the end,the algorithm employed the histogram as the feature to detect the median filtering process.The experiments show that,compared with other traditional algorithms,the proposed algorithm can not only reduce the computation time,but also acquire a higher detection rate with low dimension.