Main Article Content
Abstract
Physical characteristics analysis of metal by analyzing metal microstructures were explored to determine
the metal alloys phases in an effort to improve the mechanical characteristics of metal alloys. This research
studied the techniques for of digital image processing of metal microstructures followed by a classification
scheme phase pattern microstructures. Image pre-processings preceded the rests to reduce normally inherent
noise and to enhance get better the specific features. The extraction of the phase patterns were based on a
boundary detecting masks and a technique of threshold segmentation of RGB (red, green, blue), luminans, and
histogram spreads. The final steps of the phase pattern classification resorted to some autocorrelation methods
based on the eigen values as the distinguishing parameters. The results worked satisfactorily indicate that the
overall scheme. Quantitatively the nodular cast iron had longest randomness pattern with of a ratio of 13,18 and
high carbon steel had shortest randomness pattern with a ratio of 13,72. The percentages of the image phases
pattern widths of metal microstructures could be determined as well.
Keywords: Digital image preprocessing, metals microstructures, boundary detect segmentation, histogram
segmentation, autocorrelation.
the metal alloys phases in an effort to improve the mechanical characteristics of metal alloys. This research
studied the techniques for of digital image processing of metal microstructures followed by a classification
scheme phase pattern microstructures. Image pre-processings preceded the rests to reduce normally inherent
noise and to enhance get better the specific features. The extraction of the phase patterns were based on a
boundary detecting masks and a technique of threshold segmentation of RGB (red, green, blue), luminans, and
histogram spreads. The final steps of the phase pattern classification resorted to some autocorrelation methods
based on the eigen values as the distinguishing parameters. The results worked satisfactorily indicate that the
overall scheme. Quantitatively the nodular cast iron had longest randomness pattern with of a ratio of 13,18 and
high carbon steel had shortest randomness pattern with a ratio of 13,72. The percentages of the image phases
pattern widths of metal microstructures could be determined as well.
Keywords: Digital image preprocessing, metals microstructures, boundary detect segmentation, histogram
segmentation, autocorrelation.