Calculate the auditory masking curve of each frame of data and the minimum masking threshold of 24 subbands, and then compare it with the corresponding subband energy. When the sub-band energy involved in the audio watermarking algorithm is lower than the threshold, hidden information can be embedded in it. Statistics show that the auditory masking curve of each frame of data has the same changing law, which is approximately Gaussian in the low frequency part, and the variance is large
正在翻译中..