JPEG图像的块效应是判断图像是否被篡改的一个重要线索。本文基于对JPEG图像整体块效应的分析,定义了新的针对图像局部区域的块效应评价,并由此提出了一种有效的JPEG伪造图像盲取证方法。首先要获取待测图像在水平方向和垂直方向的差分图像,然后将两个不同方向的差分图像分别进行特定大小的分块处理,再计算每个分块区域的局部块效应评价,最后根据待测图像在不同区域局部块效应评价的明显差异分别从水平方向和垂直方向检测出图像被篡改区域的具体位置。实验结果表明,该方法可以有效的检测出经过JPEG双压缩的伪造图像,本文又进一步使用不同大小的滑块和不同的评价公式对伪造图像进行检测,并从抗干扰性和算法效率上比较不同算法的利弊,为寻找更优的JPEG伪造图像盲取证方法奠定了基础。
The blocking artifact in the JPEG images is becoming a more and more important clue in detecting image forgeries. This paper presents a new block artifact measure of the images local area and an effective method of blind forensic of JPEG forgeries on the basis of analyzing the overall block artifact of the JPEG images.Firstly,we calculate the horizontal and vertical differences maps of the JPEG image,and then divide the two maps into many sub-blocks with a specific size.Secondly,we measure the block artifact of each sub-block on the local area using the new method.Finally,respectively from the horizontal and vertical direction,the detection and location of doctored region is realized via checking significant differences of block artifact in different local areas.The experimental results validate the proposed approach in blind detecting the tampered area of the double JPEG compressed images.The slider blocks of varied sizes and different evaluation formulas are further used to detect the image forgeries.At the end of the paper,we give the pros and cons of different algorithms on the analysis of counter-interference and algorithm efficiency,which lays the foundation for further research in the area of JPEG forgeries blind forensic.