Universität Wien FIND

260068 VU Scientific image processing (2022W)

5.00 ECTS (3.00 SWS), SPL 26 - Physik
Prüfungsimmanente Lehrveranstaltung


Hinweis: Ihr Anmeldezeitpunkt innerhalb der Frist hat keine Auswirkungen auf die Platzvergabe (kein "first come, first served").


max. 15 Teilnehmer*innen
Sprache: Englisch


Termine (iCal) - nächster Termin ist mit N markiert

Dienstag 11.10. 09:45 - 12:15 PC-Seminarraum 3, Kolingasse 14-16, OG02
Dienstag 18.10. 09:45 - 12:15 PC-Seminarraum 3, Kolingasse 14-16, OG02
Dienstag 25.10. 09:45 - 12:15 PC-Seminarraum 3, Kolingasse 14-16, OG02
Dienstag 08.11. 09:45 - 12:15 PC-Seminarraum 3, Kolingasse 14-16, OG02
Dienstag 15.11. 09:45 - 12:15 PC-Seminarraum 3, Kolingasse 14-16, OG02
Dienstag 22.11. 09:45 - 12:15 PC-Seminarraum 3, Kolingasse 14-16, OG02
Dienstag 29.11. 09:45 - 12:15 PC-Seminarraum 3, Kolingasse 14-16, OG02
Dienstag 06.12. 09:45 - 12:15 PC-Seminarraum 3, Kolingasse 14-16, OG02
Dienstag 10.01. 09:45 - 12:15 PC-Seminarraum 3, Kolingasse 14-16, OG02
Dienstag 17.01. 09:45 - 12:15 PC-Seminarraum 3, Kolingasse 14-16, OG02
Dienstag 24.01. 09:45 - 12:15 PC-Seminarraum 3, Kolingasse 14-16, OG02


Ziele, Inhalte und Methode der Lehrveranstaltung

Most of the advanced methods in modern science are based on acquiring two-dimensional images or movies of a sample/phenomenon of scientific interest. In this course, you will learn the basics of quantitative image processing and analysis for scientific purposes to be able to autonomously design, implement and optimise simple image analysis workflows. You will learn how to squeeze out information out of your images of interest, without altering such information.

Basics of imaging. Digital images. Bit depth. Color. Lookup tables. Regions of Interest (ROI). Quantitative intensity analysis and algebraic operations with images. Image correlations, convolutions and deconvolutions. Fast Fourier transform and other image transforms. Image filtering (linear, non-linear and Fourier). Image thresholding and segmentation. Image annotation and labelling.

The course will make use of interactive lectures based on students participation, and of continuous assessment in the form of individual or group activities to be performed in class or at home. The course will be designed around the free image analysis software ImageJ/Fiji, which the participants are ideally expected to install on their own computer that they will bring to class. However, participants will be also encouraged to implement the workflows in other environments or languages (e.g. Matlab, Labview, C++, Python, Julia, R...) if they want to do so.

1) Introduction to the course: why do we care about images; what is an image; what image processing is and what is not; digital image processing; image sampling and quantisation
2a) Introduction to the course: image histogram; lookup table (LUT); image types and formats; color images; human vision: RGB space; pseudocolor images; brightness/contrast; image formats
2b) ImageJ/Fiji: native image formats; plugins; image properties; color images; menu bar; region of interest (ROI)
3) ROIs, thresholding, segmentation: ROI manager; manipulate ROI selections; selection menu; specify ROI; image segmentation and thresholding in ImageJ/Fiji; Morphological operators (binary, erode, dilate,…); particle analyser; watershedding
4) Image filtering and a taste of macros: digital images as matrices; element wise operations; the math menu in ImageJ/Fiji; running, saving and opening a Macro in ImageJ/Fiji; matrix operations; linear image operations; non-linear image operators; linear spatial filtering; correlation vs convolution; smoothing; Gaussian blur; background subtractions; filters in ImageJ/Fiji
5) Frequency filtering and deconvolution: spatial domain filtering; separable kernels; frequency domain filtering; Fourier series and transform; the convolution theorem; Fourier transform of a sampled function; sampling theorem; Discrete Fourier transform (DFT); Fast Fourier transform (FFT); Filtering: spatial domain vs FFT-based frequency domain
6) More on filtering: FFT low and high pass; FFT menu in Fiji; filtering out unwanted frequencies; installing useful ImageJ/Fiji plugins for filtering and deconvolution; salt and pepper noise; median filter
7) From pixel to voxels: temporal acquisitions (time lapse); image stacks in ImageJ/Fiji (average intensity and profile fitting, bleach correction, stack manipulation, image calculator, montage, 3D FFT);
8) From pixels to voxels: 3D volumes; 3D volumes in ImageJ/Fiji (3D viewer, replace, dynamic replica, 3D project, 3D script, 3D image suite, Volume Viewer)
9 ) Macros in Fiji and the Macro Recorder
10) Characterizing object dynamics in direct space: manual tracking vs automatic tracking; Particle Tracker
11) Image registration: geometric image transformations (translation, scaling, rotation,…); simple geometric mappings; homogeneous coordinates; affine transformations; combination of simple geometrical mappings; quantifying the similarity between two images; registration in ImageJ/Fiji
12) Introduction to Image Processing by “democratic” Deep Learning

Art der Leistungskontrolle und erlaubte Hilfsmittel

Assessment will be performed by means of a combination of weekly in-class and/or homework assignments (50%), plus a final exam (50%).

Mindestanforderungen und Beurteilungsmaßstab

After the assignments and the final exam, a maximum of 100 points will be reachable. 50 points or more are needed for a passing grade, determined as follows:
≥ 50 points: 4
≥ 60 points: 3
≥ 75 points: 2
≥ 90 points: 1


Content of the lecture


Zuordnung im Vorlesungsverzeichnis


Letzte Änderung: Di 11.10.2022 11:10