The Research On Information Hiding In Variable-Bit-Rate Digital Audio

With the digital multimedia and network technology developing rapidly, the communication between people has become more and more convenient and efficient, while the information security issues also have become increasingly prominent. The piracy behaviors in software, image, video and audio are seriously impacting on the market order. Even worse, in order to achieve a certain purpose, other malicious operations such as attack, tampering, forgery and so on, make the truthfulness and completeness of the contents threatened seriously. In this context, the fast development of information hiding technology has become a secret communication and an effective method for copyright protection. It can effectively curb the illegal use of digital multimedia. In the application field, audio signal in compressed format is used broadly, which is divided into CBR(Constants Bit Rate) and VBR mode(Variable Bit Rate). With a smaller file size and higher sound quality, VBR-encoded audios for compressed format receive a growing proportion in digital audio market. Therefore, researching the digital audio information hiding algorithm has high practical significance and applied prospect under VBR mode. In the thesis, we study on VBR-mp3 and VBR-aac audio. The work mainly includes the following several parts.1. The VBR-mp3 and VBR-aac audio codec theory, which mainly focuses on analysis of the key modules including quantization and Huffman coding, is analysed, in order to understand the reason, why VBR audios have better sound qulity and smaller file size in the same ratio with CBR audios.2. Current available compressed audio information hiding algorithms are summarized. What’s more, three important evaluation methods are introduced for audio information hiding, which are embedded capacity, imperceptibility and security, as well as the relationship between them.3. Proposed two information hiding algorithms for VBR MP3 s. The first method is to hide secret data by modifying the bitrate_index, while the second one makes use of the redundant space in the length of scale factor and inserts data by swapping the corresponding index with the principle of mapping, namely scalefac_compress. All the process of the two methods is implemented after the quantization. So they are simple and have a good real-time performance. Due to the advantage of VBR MP3 s standard, they also have good imperceptibility.4. Proposed an information hiding algorithm of high capacity based on VBR AACs. Combining the(1, 3, 2) matrix encoding, this method embeds data into the encoding parameter – scale factor during the encoding process. Experimental results indicate that the proposed method can embed data with high capacity and effectively guarantee the sound qulity and security.