Audio watermark embedding and extracting method based on low-frequency carrier

The invention relates to an audio watermark embedding and extracting method based on a low-frequency carrier, and belongs to the technical field of audio processing. Firstly, a binary watermark image is converted into a one-dimensional array, and then the one-dimensional array and a secret key with the same length are subjected to XOR operation to improve the security of the digital audio watermark. Secondly, generating sine wave signals with the same number of sample points according to the number of sample points of the original audio signals, then converting symbols of sample values of each half cycle of sine waves according to the watermark information, and finally superposing the converted signals and the original audio signals to obtain audio signals containing the watermark information. When the watermark is extracted, Fourier transform is carried out on an audio signal containing the watermark, a sample point with the frequency larger than five times of sine wave frequency after Fourier transform is s