Irene Vecchio, Katja Schladitz, Michael Godehardt, Markus J. Heneka

• Continuously improving imaging technologies allow to capture the complex spatial geometry of particles. Consequently, methods to characterize their three dimensional shapes must become more sophisticated, too. Our contribution to the geometric analysis of particles based on 3d image data is to unambiguously generalize size and shape descriptors used in 2d particle analysis to the spatial setting. While being defined and meaningful for arbitrary particles, the characteristics were actually selected motivated by the application to technical cleanliness. Residual dirt particles can seriously harm mechanical components in vehicles, machines, or medical instruments. 3d geometric characterization based on micro-computed tomography allows to detect dangerous particles reliably and with high throughput. It thus enables intervention within the production line. Analogously to the commonly agreed standards for the two dimensional case, we show how to classify 3d particles as granules, chips and fibers on the basis of the chosen characteristics. The application to 3d image data of dirt particles is demonstrated.