Voiceprint characteristic and audio watermark combined audio signal recognition method

摘要: The invention discloses a voiceprint characteristic and audio watermark combined signal recognition method solves problems of inaccuracy in recognition, narrow application range, digital copyright protection video contents the like prior art. includes: (1) providing sample source; (2) extracting from files at step to obtain corresponding frequency spectrum data; (3) characteristics; (4) recognizing characteristics. can further include generation addition according requirements. By combination characteristics watermarks, voice be well recognized broadcasting, live telecasting, record broadcasting conferences tests practical environment radio television station.