Article ID Journal Published Year Pages File Type
411557 Neurocomputing 2016 7 Pages PDF
Abstract

This paper proposes a new data hiding method for H.264 without intra-frame distortion drift. The embedded secret data is first distributed into n sub-secrets with matrix equation by using Shamir׳s (t, n)-threshold secret sharing. Then we choose the block with prediction mode and embed the embedded data into discrete cosine transform (DCT) coefficients of the selected block. Finally, we recover the original video as much as possible when the hidden data is extracted out by using the t sub-secrets of the video frames. The experimental results show that this new robust data hiding method not only can get more robustness, effectively avert intra-frame distortion drift and get good visual quality, but also can protect the original video.

Related Topics
Physical Sciences and Engineering Computer Science Artificial Intelligence
Authors
, , , , , ,