Acoustic Signal Compression Research Based on Wavelet and Compressed Sensing

Abstract

Due to the ocean observation and detection requirements, as well as the construction of ocean defense system engineering, collecting and transmitting a large number of acoustic signals are prerequisite for subsequent processing and analysis. Massive detection data bring a big challenge for transmission lines and underwater processors. The detection acoustic signals can be compressed for storage and reconstructed for processing. In this paper, data compression methods are proposed based on wavelet transform and compressed sensing, the two algorithms are applied and compared for acoustic signal compression. The empirical mode decomposition algorithm is applied for pro-processing. Multilayer wavelet compression is used for getting a compressed acoustic signals. Gradient projection is used for sparse reconstruction. The experimental results show that both methods can perform well in acoustic signal compression, retain the main characteristics of the original acoustic signal, and greatly reduce signals to be transmitted and processed. Wavelet compression is faster in calculation. Compressed sensing is able to achieve a higher compression ratio.