Combined audio and video watermarking using mel-frequency cepstra

Digital watermarking is a promising technique to help protect the data security and intellectual property right. In this paper a combined audio and video watermarking technique is proposed. Content-dependent information is extracted from the audio signal. The extraction uses a well-known feature set for audio, the melfrequency cepstra. The information is encoded and embedded into the video. The algorithm can be applied to the content authentication of audiovisual data in journalism, commerce and law.