基于离散余弦变换的音频数字水印技术分析-analysis of audio digital watermarking technology based on discrete cosine transform.docx

II II Abstract Audio digital watermarking technology is a kind of information hidden technology, which hides the secret message into the audio signal. It is used for the audio signal’s copyright protection or secret communication without affect the audio quality of the audio signal. The main research object of this paper is the discrete cosine transform domain audio digital watermarking algorithm. The main work includes: First, this paper introduces the meaning of audio digital watermarking technology and its basic characteristics and common application areas ; describes some common audio digital watermarking algorithms; describes the basic characteristics of the sound signals and researches human auditory characteristics, researches the discrete cosine transform theory and its application in audio digital watermarking algorithm. Second, the traditional algorithm has a defect that the watermark is only embedded into the audio signal sequence in order, so this paper proposes a energy algorithm based on energy selection. The algorithm could adaptive search for the optimum watermark embedding point based on the audio signal sequence time domain segment energy, and then embed the watermark by using discrete cosine transform. The experimental simulation results illustrates the energy algorithm has better robustness than traditional algorithm and doesn’t undermine the transparency. Third, the energy algorithm has two defects would reduce algorithm’s robustness: spatial correlation between the watermark image pixel and the watermark information maybe embedded into the point where its segment energy was weak, the energy improved algorithm is purposed in this chapter. The Arnold transform is used to remove the spatial correlation. The average energy threshold of the audio signal segment is counted. The embedding coefficient is increased when segment energy is smaller than the average energy threshold. The quality of the audio signal is reduced a small amount in order to impr