基于块效应分析的JPEG图像处理历史检测  

DETECTING JPEG IMAGES PROCESSING HISTORY BASED ON BLOCKING ARTIFACTS ANALYSIS

作　　者：蔡浩[1] 钱建生[1] 李雷达[1] 

机构地区：[1]中国矿业大学信息与电气工程学院,江苏徐州221116

出　　处：《计算机应用与软件》2015年第12期165-169,共5页Computer Applications and Software

基　　金：国家自然科学基金项目(61379143;51204175);国家高技术研究发展计划项目(2012AA062103);徐州市科技计划项目(XM13B119)

摘　　要：针对JPEG图像及其块效应在空间域中产生的规律性周期模式,提出一种新的方法来分析块效应的周期性,以检测JPEG图像的处理历史。对于品质因子不同的一次压缩JPEG图像和不对齐双压缩JPEG图像来说,它们的块效应峰值能量分布是有差异的。首先构造图像的块效应模型,然后从模型的频谱中提取出峰值窗口,最后从峰值窗口中获得有代表性的特征进行SVM训练。实验结果表明,该方法对于判断一次压缩JPEG图像的压缩强度和检测不对齐双压缩JPEG图像的处理情况是有效且实用的。JPEG images and their blocking artifacts generate a regular periodic pattern in spatial domain. In light of this,we present a novel approach for analysing the periodicity of blocking artifacts so as to detect the processing history of JPEG images. For both the oncecompressed JPEG images with different quality factors and the non-aligned dual-compressed JPEG images,their peaks＇ energy distributions of blocking artifacts have difference. First,we build a model of blocking artifact,and then extract a peak window from the Fourier spectrum of the model. Finally,we derive representative features from peak window and carry out support vector machine（ SVM） training. Experimental results demonstrate that the proposed method is effective and practical in estimating the compression intensity of once-compressed JPEG images and detecting the processing situation of non-aligned dual-compressed JPEG images.

关 键 词：块效应 JPEG图像 周期性分析 压缩强度 不对齐双压缩 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]