基于SGF-IABC的JPEG图像下采样因子检测算法

Downscaling factor estimation of JPEG image based on SGF-IABC

对图像进行重采样操作是图像篡改的重要手段,JPEG图像作为最广泛的应用载体,对其进行重采样检测研究具有重要意义。JPEG图像由于重采样引起的块效应移位,差分其相邻极值间隔分布直方图呈周期性,且周期性受到图像纹理和边缘的干扰较大,为削弱下采样因子估计时易受到此类干扰,本文提出一种可抵抗图像纹理边缘的JPEG图像下采样因子估计模型。首先,将JPEG图像经过二次引导滤波(Secondary Guidance Filtering,SGF)的预处理去除图像的纹理噪声;其次,使用改进人工蜂群算法(Improved Artificial Bee Colony algorithm,IABC)对其进行边缘检测、提取和移除,弱化图像边缘周期性特征,以此减少图像极值区间直方图的统计误差,提升下采样因子估计值的准确性;最后,计算图像差分极值直方图,得到最终估计。实验结果表明,本文方法比现有算法具有更小的平均误差,对图像纹理和边缘在像素域的周期性干扰具有更好的鲁棒性,能有效抵抗纹理和边缘的双重干扰。

Image resampling is an important means of image tampering.As the most widely used carrier,it is of great significance to resample and detect JPEG images.According to the block effect shift of JPEG image caused by resampling,the histogram of the adjacent extreme interval distribution of the difference is periodic,and the periodicity is greatly disturbed by the texture edge of the image,in order to weaken this kind of interference in downscaling factor estimation.In this paper,a downscaling factor estimation model of JPEG image is proposed,which can resist the texture edge of the image.Firstly,the image is preprocessed by secondary guided filtering(Secondary Guidance Filtering,SGF)to remove the texture noise of the image.On this basis,the improved artificial bee swarm algorithm(Improved Artificial Bee Colony algorithm,IABC)is used to detect,extract and remove the edge of the image,so as to weaken the periodic features of the image edge,so as to reduce the statistical error of the image extreme interval histogram.Finally,the image difference extreme histogram is calculated to improve the accuracy of downscaling factor estimation.The experimental results show that this algorithm has smaller average error than the existing algorithms,has better robustness to the periodic interference of image texture and edge in pixel domain,and can effectively resist the double interference of texture and edge.