Automatic detection and segmentation of indentations and tribological marks by traceable machine vision system

Automatically segmenting topographical features improves precision of characterization, essential in mechanical characterizations requiring the measurement of residual marks, e.g. Vickers hardness, pin-on-disc. This work demonstrates a method developed within the 22RPT01 BVH-K project which leverages machine vision to segment Vickers indentations and wear tracks on industrially relevant applications, e.g. metals, polymers. The method exploits topographical microscopes, providing traceability, and allows higher precision than operator-based approaches. Isolation of local plasticity effect on feature edge, e.g. pile-up, galling, is also possible, thus providing a strong tool to support primary metrology in improving candidate reference materials and accuracy by correcting edge-effects.