Wavelet Denoising Preprocessing for Lessening Measurement Errors in Near Field Acoustical Holography

Abstract

Near field acoustical holography (NAH) method can construct an entire sound field by measuring sound pressure on a hologram plane. There are many sources of measurement errors when one attempts to construct a hologram of near field acoustical holography, which are enlarged in the reconstruction of the source field and impact the reconstruction results. Wavelet processing applied to pressure measurements in the near field provides a solution to lessen the effects due to measurement errors of the hologram. The approach is based on wavelet denoising algorithms. The basic principle is based on the feature of noise and signals in the measurement wavelet spectra with different performance characteristics. The wavelet threshold denoising algorithm is adopted to remove the wavelet spectra generated by noise and retain the original signal wavelet spectra. Experimental results show the validity of the wavelet denoising processing and sound source identification has achieved remarkable results.