Image Processing for Precision Atomic Force Microscopy

Atomic Force Microsopy (AFM) is an increasingly popular means of obtaining accurate metrological data of surface topography. The AFM measures surface topography by recording the vertical motion of a probe as it traces the profiled surface in a raster fashion. However, the metrological accuracy of AFM images is limited by image convolution between the probe tip and the sample surface. As a result of the development of new precision engineering applications, the critical dimensions of engineered features are increasingly pushed into the sub-pm range. To maximize the utility of the AFM in such applications, it is imperative therefore to develop strategies to minimize inaccuracies in AFM imaging. In this thesis, we examine the process of image convolution and develop a complete strategy to optimize the fidelity of AFM measurements via processing of the obtained AFM images. Through such an approach, we seek to maximize the potential of existing AFM equipment for metrological measurements. Thesis Supervisor: Kamal Youcef-Toumi Title: Professor of Mechanical Engineering Acknowledgments I would like to extend my heartfelt gratitude first to my mentor, Professor Youcef-Toumi for his invaluable guidance. In addition, I am also indebted to my colleagues in the Mechatronics Research Laboratory, especially Bernardo for their input during discussions. Credit must also go to Bae-Ian Wu for his counsel in the visual presentation of this thesis. I would also like to thank those in church, particularly James Lee and Heechin Chae who have guided and sustained me through my time in Boston. Lastly, to my parents without whom all this would not have been possible.