Olaf Kühn

Model-based edge detection in height map images with nanometer resolution

The advances at the semiconductor fabrication cause a need for three-dimensional measurements within the micro- and nanometer range. A method to derive lateral measuring information from height profiles with nanometer resolution is shortly introduced. Thereby, measurements are taken with a high resolution, point-wise working laser sensor. The paper discusses in detail the analysis of the measuring signal as observed at different height standards. At height standards with discontinuities smaller than the coherence length of the laser the measuring signal overshoots. This effect is called batwing effect. The novel approach comprises the utilisation of the batwing effect to determine the edge position. The characteristic signal curve is modelled with an appropriate mathematical function. Thus, edge detection algorithms well known from optical precision measurements are modified and applied to calculate the location of the edge along the scan line of the laser sensor. Exemplary, an experimental standard deviation of 8.2 nm of the width of a 69 nm high height standard is attained. The measurements with the laser focus sensor are compared with AFM measurements at the same height standard. The presented work results from the collaborative research centre (SFB) 622 supported by the German Research Foundation (DFG).

Adaptive Approach to the Estimation of Uncertainty in Industrial Quality Testing Using Image Sensors

Abstract For drawing the right conclusions from an indicated measuring value of a quantity to the unit under test, it is indispensable to express the assigned expanded measurement uncertainty. For the industrial quality testing image sensors are widely used due to their multitude of advantages, e. g. the contactless operation mode, the high level of automation connected with the short measurement time and the flexibility in the measurement of different workpieces. This paper is derived to propose a new knowledge-based approach to the adaptive and automated uncertainty estimation for precision coordinate measurements using image sensors. It can be shown, that the novel method gives estimates of measurement uncertainty which are practicable and widely corresponding with experimental effectuated uncertainty declarations. Zusammenfassung Damit bei der Qualitätssicherung der zu einer Messgröße angezeigte Messwert richtige Rückschlüsse auf das Messobjekt zulässt, ist eine Aussage über die zugehörige Messunsicherheit unabdingbar. Aufgrund der vielen Vorteile, wie z. B. dem berührungslosen Messprinzip, dem hohen Automatisierungsgrad verbunden mit geringen Messzeiten und der Flexibilität bei der Messung verschiedenster Werkstücke, werden Bildsensoren für die industrielle Qualitätssicherung sehr verbreitet eingesetzt. Der vorliegende Beitrag beschreibt einen neuen wissensbasierten Ansatz zur adaptiven und automatisierten Ermittlung der Messunsicherheit bei Präzisionskoordinatenmessungen mit Bildsensoren. Es wird gezeigt, dass die neue Methode Schätzwerte für die Messunsicherheit liefert, die praktikabel sind und weitgehend mit experimentell erzielten Messunsicherheitsangaben übereinstimmen.