Education

  • Ph.D., Electrical and Computer Engineering, Georgia Tech, Aug. 2014
  • M.Sc., Mathematics, Georgia Tech, May 2013
  • Diplom-Ingenieur, Engineering Cybernetics, Univ. of Stuttgart, Germany, Feb. 2010
  • M.Sc., Engineering Science and Mechanics, Georgia Tech, Dec. 2009

Publications

Journal Papers

  • M. Mueller, P. Karasev, I. Kolesov, A. Tannenbaum , Optical Flow Estimation for Flame Detection in Videos, IEEE Transactions on Image Processing, Vol. 22, No. 7, pp. 2786 - 2797, 2013
  • M. F. Müller, J.-Y. Kim, J. Qu and L. J. Jacobs, Characteristics of Second Harmonic Generation of Lamb Waves in Nonlinear Elastic Plates, Journal of the Acoustical Society of America, 2010

Conferences

  • M. Mueller, P. Karasev, I. Kolesov, A. Tannenbaum , A Video Analytics Framework for Amorphous and Unstructured Anomaly Detection, in Proceedings of IEEE International Conference on Image Processing, 2011
  • M. F. Müller, J.-Y. Kim, J. Qu and L. J. Jacobs, On the Excitability of Second Harmonic Lamb Waves in Isotropic Plates, in Proceedings of 36th Annual Review of Progress in Quantitative Non-destructive Evaluation Conference, 2009

Theses

  • M. F. Mueller, Advisor: Prof. Yezzi Physics-driven Variational Methods for Computer Vision and Shape-based Imaging, Ph.D. Thesis, Georgia Institute of Technology, Atlanta, 2014
  • M. F. Müller, Advisor: Prof. Jacobs Analytical Investigation of Internally Resonant Second Harmonic Lamb Waves in Nonlinear Elastic Isotropic Plates, Master's Thesis, Georgia Institute of Technology, Atlanta, 2009
  • M. F. Müller, Advisor: Prof. Gaul Wave Propagation in a Friction-Coupled Three-Rod System, Student Research Project, University of Stuttgart, 2008

Ph.D. Research Projects

Optical Flow for Vision-Based Flame Detection

Fire motion is estimated using optimal mass transport optical flow, whose motion model is inspired by the physical law of mass conservation, a governing equation for fire dynamics. The estimated motion fields are used to first detect candidate regions characterized by high motion activity, which are then tracked over time using active contours. To classify candidate regions, a neural net is trained on a set of novel motion features, which are extracted from optical flow fields of candidate regions.

Coupled Photo-Geometric Object Features

Active contour models for segmentation in thermal videos are presented, which generalize the well-known Mumford-Shah functional. The diffusive nature of heat processes in thermal imagery motivates the use of Mumford-Shah-type smooth approximations for the image radiance. Mumford-Shah's isotropic smoothness constraint is generalized to anisotropic diffusion in this research, where the image gradient is decomposed into components parallel and perpendicular to level set curves describing the object's boundary contour. In a limiting case, this anisotropic Mumford-Shah segmentation energy yields a one-dimensional "photo-geometric" representation of an object which is invariant to translation, rotation and scale. These properties allow the photo-geometric object representation to be efficiently used as a radiance feature; a recognition-segmentation active contour energy, whose shape and radiance follow a training model obtained by principal component analysis of a training set's shape and radiance features, is finally applied to tracking problems in thermal imagery.

Adjoint Active Contours for Shape-Based Imaging

The goal of this research is to estimate both location and shape of buried objects from surface measurements of waves scattered from the object. These objects' shapes are described by active contours: A misfit energy quantifying the discrepancy between measured and simulated wave amplitudes is minimized with respect to object shape using the adjoint state method. The minimizing active contour evolution requires numerical forward scattering solutions, which are obtained by way of the method of fundamental solutions, a meshfree collocation method. In combination with active contours being implemented as level sets, one obtains a completely meshfree algorithm; a considerable advantage over previous work in this field. With future applications in medical and geophysical imaging in mind, the method is formulated for acoustic and elastodynamic wave processes in the frequency domain.



DISCLAIMER: THIS PAGE IS NOT A PUBLICATION OF THE GEORGIA INSTITUTE OF TECHNOLOGY AND THE GEORGIA INSTITUTE OF TECHNOLOGY HAS NOT EDITED OR EXAMINED THE CONTENT. THE AUTHOR(S) OF THE PAGE ARE SOLELY RESPONSIBLE FOR THE CONTENT.