Morphological filtering via energy minimization
Master Semester Project
The advantage of morphological filters is their invariance with respect to any change of contrast; they can typically be applied to biomedical data. It was recently shown by J. Darbon that a practical and concise energy minimization formulation (total-variation with L1 data fidelity) satisfies the same property. In this project, we thus want to perform morphological denoising by minimizing this energy. An efficient implementation which uses the problem structure will be developed. Exploiting the solution properties, specific examples should then show the benefits of the method. Requisites : courses in signal/image processing, interest for algorithmic methods and general knowledge in programming (MATLAB and/or Java).
- Michael Unser, email@example.com, 021 693 51 75, BM 4.136