Font Size: a A A

Study On Information Hiding Algorithm Based On Prediction Modes For H.264Video

Posted on:2015-02-12Degree:MasterType:Thesis
Country:ChinaCandidate:J M LiuFull Text:PDF
GTID:2268330428975993Subject:Information security
Abstract/Summary:PDF Full Text Request
In recent years, people use information hiding technology in images, audio, video and other carrier to realize covert communication of confidential information. Based on the carrier itself information redundancy, video information hiding technology that takes multimedia video as carriers, can be better hidden the existence of secret information.In this thesis, the video information hiding algorithm is based on the H.264codec. Two video information hiding schemes have been designed based on intra-frame prediction of H.264in this thesis. The main work is as follows:(1) A video information hiding algorithm based on LSBMR has been designed. When video luminance is intra predicted, there are several areas whose luminance intra-prediction is independent of each other in the macroblock coded by I4MB mode. In related areas,2luminance prediction modes are selected for information embedding, and LSBMR information hiding algorithm is used to embed the secret information by modulating the selected prediction modes. In order to avoid an excessive impact on the video quality, an intra-prediction mode mapping rule is proposed to select less costly suboptimal mode. Experimental results show that test video has still good objective and subjective information quality after embedding information.(2) Combining with matrix encoding, two video information hiding algorithms have been designed:hamming+1video information hiding algorithm and high payload (7,4) hamming video information hiding algorithm. Two algorithms are realized information hiding in the intra-prediction process.Hamming+1algorithm combines the matrix encoding and LSBMR algorithm. In this scheme,8luminance prediction modes is selected for information embedding. First, the matrix encoding is used to embed3bits secret information into7prediction modes, and then LSBMR algorithm is introduced by analyzing sub-LSB or LSB of the8prediction modes to achieve4bits secret information embedding at last.High payload (7,4) Hamming code algorithm is based on matrix encoding,(7,4) Hamming code is used to cut0to27into16categories, and each category includes8bits vector of length7. In this scheme,7luminance prediction modes are selected for information embedding. When the secret information is embedded into the cover video, the7bits secret information is divided into the information bit data and check bit data and then the bit vector related to check bit data is found in the category associated with the information bit data, and the LSB of the7prediction modes are replaced to embed the secret information. Experimental results show that the two algorithms are available to dramatically increase the capacity of information hiding without compromising video quality.
Keywords/Search Tags:Video information hiding, Intra-frame prediction, LSBMR, Hamming+1, Highpayload (7,4) hamming
PDF Full Text Request
Related items