Wavelet denoising to improve the quality of SEM images

Scanning Electron Microscopy (SEM) is a useful and mighty powerful technique in diverse science fields.An acute problem plaguing the image inspection of soft organic and polymeric samples is the mecanical and thermal damage of the sample by electron bombardment. To preserve the sample integrity a fast scanning is the usual solution, but this mean a loss of image quality. Here we show how the application of wavelet denoising as a post-processing step yields images with comparable quality to those obtained with a slow scanning speed. Specifically, the the linear shift invariant discrete wavelet transform provided the best perceptual quality.

keywords: Scanning Electron Microscopy (SEM), Discrete Wavelet Transform (DWT), image denoising, image enhacement