Broadcast Audio Signal Similarity Detection Method Based on Subband Processing and Volterra Adaptive Filtering

Due to the wide audience and great influence of radio programs, its broadcast control requirements were high, and the tolerance for abnormal broadcasts such as mis-broadcasting, interrupted broadcasts, and missed broadcasts was low. Aiming at the real-time monitoring of broadcast programs, this paper proposed a fast broadcast audio signal similarity detection method. This method calculated the Pearson correlation coefficient to determine whether the two broadcast audio signals were similar. Then, in order to offset the influence of the codec, transceiver equipment and transmission channel, an adaptive Volterra filter was applied to process the signal. Finally, the sub-band decomposition technology was used to decompose the full-band signal into sub-band signals, and only the sub-band with the highest power was analyzed and pre-processed to reduce the amount of calculation. The experimental results show that, after considering the real transmission impact and comparing different conditions, the method has good detection accuracy and a small amount of calculation, which can meet the requirements of real-time processing.