Design And System Implementation Of An Audio Watermarking Algorithm Based On DFT And DWT

With the development of network technology and digital multimedia technology, digital watermak has become a popular topic in the field of digital copyright protection. Audio digital watermarking technique is a kind of digital watermarking that takes the audio signal as carrier. With embedding data into the audio signal, it implements some goals such as copyright protection and cover communication.This thesis mainly focuses on the problem of embedding satisfactory robust and im- perceptible watermarking into the audio digital products. Main idea as follows:First, after introducing general conception of digital watermarking, this thesis implements two audio watermarking algorithm. One is based on discrete Fourier transfer, which embeds the watermark into the amplitude of the result of DFT by using quantization method. The experiments show that the algorithm has good imperceptible and some robustness. The other is based on discrete wavelet transfer, which embeds the data into the approximation coefficients by using quantization method. And it also has imperceptible and robustness to many signal operation.After discussing the above algorithm, we proposed a new algorithm based on both DWT and DFT. It transfers the audio signal using DWT fistly and then choose the lowest frequency coefficient to DFT. Then it embeds data into amplitude coefficient. The experiment result indicates that the new algorithm is inaudible and robust against many signal operation. In addition, comparing with the two algorithms above under the same condition, it proves that the new algorithm is better than the other two.Lastly, we design and implement an audio application system using Matlab tool based on the discussion of the three algorithms.