Beneficial Image Preprocessing by Contrast Enhancement Technique for SEM Images

Abstract

In this paper a morphological filtering algorithm using an exposure thresholding and measures of central tendency hasbeen proposed for solving the low contrast of Scanning Electron Microscopic (SEM) images of composite materials foraccurate Filler Content Estimation. SEM image of a composite material comprises visible morphological structures likefillers such as silica nanoparticles. The SEM image analysis via segmentation will assist in the study of distribution of thesestructures. The estimation of the filler content is more accurate only when the SEM images have proper contrast for analysisif not the results lead to less accuracy. To overcome this drawback, we have proposed a preprocessing technique to increasethe contrast of SEM images. So that the preprocessed image can be used for post processing namely segmentation and hencethe error is less for filler content estimation. We introduced the transformations using morphological processing to extractthe bright and darker features of the images. The optimum threshold value is determined by the image exposure. A detailedcomparative analysis with other existing techniques has been performed to prove the superior performance of the proposedmethod.