Digital image splicing detection based on Markov features in QDCT and QWT domain

摘要

Image splicing detection is of fundamental importance in digital forensicsand therefore has attracted increasing attention recently. In this paper, acolor image splicing detection approach is proposed based on Markov transitionprobability of quaternion component separation in quaternion discrete cosinetransform (QDCT) domain and quaternion wavelet transform (QWT) domain. Firstly,Markov features of the intra-block and inter-block between block QDCTcoefficients are obtained from the real part and three imaginary parts of QDCTcoefficients respectively. Then, additional Markov features are extracted fromluminance (Y) channel in quaternion wavelet transform domain to characterizethe dependency of position among quaternion wavelet subband coefficients.Finally, ensemble classifier (EC) is exploited to classify the spliced andauthentic color images. The experiment results demonstrate that the proposedapproach can outperforms some state-of-the-art methods.