Beschreibung:
AbstractThis paper describes an effective and simple procedure to derive information about the dislocation density distribution in metals by applying standard techniques of digital image processing on gray scaled microstructure images obtained from transmission‐electron microscopy. In a representative transmission‐electron microscopy image, two local dislocation density values were investigated by classical methods and were used as input parameter for further processing. A correlation between dislocations and image intensities is assumed such that dark areas in microstructural images are seen as a dense concentration of dislocations. Then, the contrast is increased for each transmission‐electron microscopy photography. In the next step, posterization, a gradation of tone, is applied to these images. From this, a pixel weighted average distribution of gray level correlated dislocation densities is obtained as well as an average value for a given set of images. The implementation of the several processing steps is done in Matlab employing graphical user interfaces.