SOME NOTES ON THE APPLICATION OF DISCRETE WAVELET TRANSFORM IN IMAGE PROCESSING

Mathematical transforms are used in signal processing in order to extract what is known as “hidden” information. One of these mathematical tools is the Discrete Wavelet Transform (DWT), which has been increasingly employed in non‐destructive testing and, more specifically, in image processing. The main concern in the present work is to employ DWT to suppress noise without losing relevant image features. However, some aspects must be taken into consideration when applying DWT in image processing, mainly in the case of weld radiographs, in order to achieve consistent results. Three topics were selected as representative of these difficulties, as follows: 1) How can image matrix be filled to fit the 2n lines and 2n rows requirement? 2) How can the most suitable decomposition level of the DWT function and the correct choice of their coefficient suppression be selected? 3) Is there any influence of the scanning direction and the weld radiograph image, e.g., longitudinal or transversal, on the final processing ...