作者: Su Yu , Hui-Xiang Xu , Zhi-Hong Yang , Zhi-Chao Li
关键词: Interpolation 、 Detection theory 、 Image (mathematics) 、 Computer vision 、 Steganalysis 、 Artificial intelligence 、 Motion estimation 、 Mathematics 、 Motion interpolation 、 Intrusion detection system 、 Steganography
摘要: This paper studies the potential for passive steganalysis in correlated image frames using non-classical detection theory.In particular,an algorithm digital video steganalysis,named MoViSteg Motion-based Video Steganalysis, is developed that exploits temporal correlation among individual signals to enhance performance. The method differs from prior art use of motion interpolation and asymptotic memoryless we believe well-suited steganalysis. Results discussion are provide order demonstrate our ideas intrusion a broad class emerging multimedia applications.