Grain Size in Martensitic Steel

One martensitic stainless steel image is submitted for analysis. The absence of clear grain boundaries makes martensitic structures difficult to segment using traditional threshold-based methods.  A machine-learning algorithm specifically developed for martensitic steel allows a reproducible one-step detection.

Figure 1. Image of martensitic stainless steel at 100X.

Figure 2. Grain boundaries are detected using the Martensitic algorithm prior to measurement.


Reproducible detection and measurement of grain size as per ASTM E112 was performed using Clemex Vision image analysis software calibrated with a NIST-traceable stage micrometer.


ASTM E112 grain size measurements are performed. Results are cumulated for automated statistics and graph generation. Final results can be printed directly from Clemex Vision. Raw data are linked to their respective objects and can be exported in Excel format.