<![CDATA[Physical characteristics analysis of metal by analyzing metal microstructures were explored to determinethe metal alloys phases in an effort to improve the mechanical characteristics of metal alloys. This researchstudied the techniques for of digital image processing of metal microstructures followed by a classificationscheme phase pattern microstructures. Image pre-processings preceded the rests to reduce normally inherentnoise and to enhance get better the specific features. The extraction of the phase patterns were based on aboundary detecting masks and a technique of threshold segmentation of RGB (red, green, blue), luminans, andhistogram spreads. The final steps of the phase pattern classification resorted to some autocorrelation methodsbased on the eigen values as the distinguishing parameters. The results worked satisfactorily indicate that theoverall scheme. Quantitatively the nodular cast iron had longest randomness pattern with of a ratio of 13,18 andhigh carbon steel had shortest randomness pattern with a ratio of 13,72. The percentages of the image phasespattern widths of metal microstructures could be determined as well.Keywords: Digital image preprocessing, metals microstructures, boundary detect segmentation, histogramsegmentation, autocorrelation.]]>
Copyrights © 2007
