Detection of double compression in HEVC videos based on the statistical characteristic of DCT coefficient
Meilin Huang · Journal of Optoelectronics·laser · 2015
With the wide spreading of powerful and easy-to-use video editing software,digital videos are exposed to lots of different tampering kinds which destroy the authenticity and integrity of videos.Since double compression is necessary to video tampering,the detection of double compression is an important supplementary means for digital video forensics.By analyzing changes in discrete cosine transform(DCT)coefficients caused by quantization,an algorithm of detecting double compression for high efficiency video coding(HEVC)videos in different quantization parameters is proposed,which employs 22-dimensional union feature set consisting of the DCT histogram and the parity group statistics of DCT coefficient-pair,and then the feature set is sent to support vector machine(SVM)classification to distinguish double compression videos from single compression videos.Experimental verify show the effectiveness of the algorithm.