General Information
- Lecturer: Prof. Dr. Fred Hamprecht
- TA: Bjoern Andres
- Thursday, 14.00-18.00
- INF 350 OMZ R U012
Contents
Modern imaging devices have a field of view that ranges from microscopic to galactic, with applications ranging from biology to medicine to industry to astronomy. While the physical realizations of these instruments differ widely, the underlying principles are the same; the aim of this lecture is to convey underlying principles such as
- Fourier optics
- Fourier Slice Theorems
- Interferometry
- Holography
02.04.2009 - Introduction, Tomographic Projections, Sinograms
- Overview of Imaging Techniques And Their Physical Background
- Tomographic Projections, Sinograms And Their Application In CT Imaging
- Exercise 1: Tomographic Projections, Sinograms
- Literature: [Goodman 2006]
09.04.2009 - Ridge Regression, Algebraic Reconstruction from Tomographic Projections
- Ridge Regression as an Optimization Problem, Computation by Means of the Pseudo-Inverse Matrix, Statistical Interpretation.
- The Projection Slice Theorem and Its Application to Image Reconstruction (intuitively)
- Ram-Lak and Hamming Filtering
- Exercise 2: Ridge Regression, Algebraic Reconstruction, Ram-Lak and Hamming Filtering
- Literature: [Hastie 2009]
16.04.2009 - Projection Slice Theorem, Fourier Reconstruction
- The projection slice theorem and its application to image reconstruction (rigorously)
- Exercise 3: Fourier Reconstruction
- Literature: [Goodman 2006]
23.04.2009 - Scalar Diffraction Theory
- From the Maxwell Equations To The Kirchhoff Integral
- Huygens–Fresnel Principle
- Fresnel Approximation
- Fraunhofer Approximation
- Exercise 4: Scalar Diffraction Theory
- Literature: [Goodman 2006]
31.04.2009 - Ray Optics, Phase Contrast Microscopy
- Deriving Ray Optics From Fourier Optics
- Phase Contrast Microscopy
- Literature: [Goodman 2006]
07.05.2009 - Microscopy Beyond the Diffraction Limit
- Exercise 5: Super-Resolution of STORM Imagery
- Literature: [Hell 2007, Hell 2009]
14.05.2009 - Holography
- Literature: [Lauterborn 2003]
28.05.2009 - Computational Photography
- Trade-offs of Light-Field Cameras
- Plenoptic Camera
- Coded Aperture Photography
- Exercise 6: Coded aperture imaging
- Literature: [Levin et al. 2008, Ng et al. 2005]
04.06.2009 - Time-of-Flight Range Sensors
- Principles of Time-Of-Flight Range Sensors
- Depth Map Denoising
- Normalized Convolution
- Bilateral Filtering
- Exercise 7: Time-of-Flight Range Sensors
18.06.2009 - Intravascular Ultrasound
- Hilbert Transform
- Intravascular Ultrasound
- Exercise 8: Intravascular Ultrasound
- Literature: [Hoskins 2002]
25.06.2009 - White Light Interferometry
02.07.2009 - Magnetic Resonance Imaging (MRI)
- Nuclear Magnetic Resonance (NMR)
- Spatial Encoding: Slice Selection, Frequency Encoding, Phase Encoding
- Gradient De-phasing and Re-phasing
- Gradient Echoes And Spin Echoes
- Exercise 10: Magnetic Resonance Imaging
- Literature: [Mitchell 1999]
09.07.2009 - Magnetic Resonance Imaging (MRI)
- Magnetic Resonance Spectroscopic Imaging (MRSI)
- Parallel MRI (SENSE)
- Other modalities: fMRI, DTI
- Literature: Cf. http://www.mr.ethz.ch/sense/, [Mitchell 1999]
Recommended Literature
- [Goodman 2006]. Introduction to Fourier Optics.
- [Lauterborn 2003]. Coherent Optics.
- [Oppenheim 1989]. Discrete-Time Signal Processing.
- [Mitchell 1999]. MRI Principles.
- [Hoskins 2002]. Diagnostic Ultrasound.
- [Hastie 2009]. The elements of statistical learning.
Regulations
We pose MATLAB programming exercises which are usually started in the last hour of the allotted lecture time and should be completed as homework. The completed solutions (code and plots in hardcopy) will always be collected at the start of the next lecture session. If there are problems with printing, the solutions may also be sent to the e-mail address listed below: in this case, please enclose everything into a single PDF document (e.g. use the publish command of MATLAB for creating a LaTeX document and run latex, dvips and ps2pdf).
There will be 15 exercise sheets. The ECTS grade on your course certificate will depend on your success in these exercises (i.e. there will be no additional examination).
Please hand in the MATLAB code, figures and explanations (describing clearly which belongs to which). Non-trivial sections of your code should be explained with short comments, and variables should have self-explanatory names. Plots should have informative axis labels, legends and captions.
The names of the most relevant new MATLAB commands are usually started at the end of the exercise. You may hand in the exercises in teams of two people, which must be clearly named on the solution sheet. Discussions between different teams about the exercises are encouraged, but the code must not be copied verbatim (the same holds for any implementations which may be available on the WWW). Please respect particularly this rule, otherwise we cannot give you a passing grade.
If you plan to elect the course for your master's or diploma finals (as minor "Computer Science" in combination with another course), an additional oral examination about the contents of the lecture is required: in this case, please arrange a date with Prof. Hamprecht.
Please pose any questions concerning the exercises preferably via e-mail (Send Mail)
or by phone (06221/545280).