Grain Size in Austenitic Steel
One austenitic stainless steel image is submitted for analysis. Etch stains, polishing artifacts, and carbide precipitates in the grain or at grain edges make traditional thresholding inconsistent. A machine-learning algorithm specifically developed for austenitic grains allows a reliable one-step segmentation.
Figure 1. Image of austenitic stainless steel at 100X.
Figure 2. Grain boundary is segmented using the Austenitic algorithm prior to measurement.