Comparative study of noise reduction in ultrasonic inspection system

The effective noise reduction schemes in ultrasonic inspection have shown significant role for detection of flaws in materials. This paper presents the comparative result of noise reducing schemes based on Hilbert-Huang transform and Wavelet transform. The basic principle of HHT scheme includes two parts which are empirical mode decomposition (EMD) algorithm and a sum of intrinsic mode functions (IMF). By doing EMD process, the valuable information out of individual IMF can be maintained with reduced noise level for feature vectors. In the Wavelet transform, the signals are decomposed into low and high information and the feature vectors can be selected by Wavelet coefficients. In order to compare the performance of the two distinct schemes, this paper utilizes the soft and hard thresholding criterion on WT and HHT. Key-Words: Ultrasonic inspection, Automatic signal processing, Hilbert-Huang Transform, Wavelet Transform, Denoising scheme, Nondestructive evaluation.