An automatic method of measuring the cristallographic orientations of Fe-Si alloys using electron scan microscope image processing and etch pitting

Abstract
We propose a new method of measuring the crystallographic orientations for metallurgical applications. This method involves measuring these orientations by electron scan microscope image computing. After image segmentation obtained through the use of an algorithm that makes an area approach and an edge detector using spline functions work together, we classify these areas in three different categories. To separate two of them, we have to make an edge detection using multi-scale laplacian, and a skeletonization. This latter operation is computed with a two-step thinning algorithm, with different criteria to prevent the formation of unexpected segments. At the end, a specific measure was made for each of the three kinds of images, in order to extract the crystallographic orientations of any textured iron sheet.