Marker-free identification of milled surfaces by analyzing stochastic and kinematic surface features by means of wavelet transformation

This article presents a marker-free component identification of milled workpiece surfaces. For the identification, unique features from a 2-D profile are detected in the 3-D frequency domain. Knowing the influence of process parameters, tool runout, cutting edge roughness during flank milling on the stochastic surface generation, a profile cut can be set, which is as unique as a human fingerprint. Experimental investigations show a false-positive-rate of 10-20 for a confocal measurement of the surface as well as for the usage of an industrial camera.