A Wavelet Threshold Denoising Method for Ultrasonic Signal Based on EMD and Correlation Coefficient Analysis

Mingjian Sun,Yi Shen,Wei Zhang
DOI: https://doi.org/10.1109/cisp.2010.5647811
2010-01-01
Abstract:Ultrasonic image processing apparatus usually use TGC amplifiers for the appropriate amplification of far-field echoes to compensate for image quality reduction due to far-field signal attenuation. However, along with the signal enhancement, the far field noises are also amplified in a certain level. For a scanning line of RF, it generates a type of noise with large far field noises and small near field noises, which is named the trumpet type noise. In this paper, the trumpet type noise is studied and a denoising method is proposed based on EMD (Empirical mode decomposition) and correlation coefficients analysis. By applying correlation analysis to two consecutive frame of echo signal, regions of different intensity of the noise are dealt with by threshold after being distinguished. Experimental results demonstrate its applicability to denosing.
What problem does this paper attempt to address?