Image Processing School

Application review process:

Student applications will be resolved by Friday, August 7.

Course description:

We will teach in the context of Fiji (http://pacific.mpi-cbg.de ), an open source software application with ImageJ2 at its core. Fiji comes with online wiki documentation and a set of tutorials, and is fully expandable by plugins written in a variety of languages (java, jython, javascript, jruby, clojure, beanshell and its own ImageJ macro language). Dashes of Blender (http://www.blender.org ) as well for movie composition and 3D mesh editing and visualization.

1. Fundamentals of Image Processing:

  • imageAnatomy of an image (channels, bit depth, lookup tables)
  • Image generation (light band-pass lters, long-pass lters, channel assignment)
  • Image le formats and their usages (lossy, non-lossy, compression, metadata)
  • Key concepts for optimal image acquisition at confocal and at TEM (dynamic range, optics setup)

2. Image analysis:

  • Colocalization analysis.
  • Calibration.
  • Measurements of pixel intensity and object dimensions.

3. Image registration:

  • Feature extraction for automatic registration.
  • Rigid registration: aligning stack slices, overlaying images.
  • Non-rigid/elastic registration: usages in transmission electron microscopy.

4. Image segmentation and 3D modeling:

  • Automatic segmentation: neurite tracers, growing areas, active contours, level sets.
  • Manual segmentation.
  • Generation of 3D models from segmentations.

5. Batch image processing and analysis:

  • High-level (easy) programming for automatic processing and measurement of large image col-
  • lections.
  • Hardware setup tips for increased performance.
  • Strategies for using computer clusters and parallelization in general.

6. Pixel processing and analysis:

  • Low-level programming for pixel processing, editing and measurement.
  • Image creation, retrieval and storage.

Contact

No longer applicable

Target group

Neuroscience graduate students and postdocs who need to master image processing skills, from image basics to task automation with scripts and custom software.

Useful background knowledge

Familiarity with optical and electron microscopy images will help, but it's not necessary.

Why are we organizing this course?

Our motivation stems from two main sources: on the one hand, the realization that every day work in all aspects of neuroscience involve image acquisition and processing; on the other hand, the fact that those engaged in image processing activities are generally very poorly educated in the matter, or not at all. We would like to bridge this knowledge gap. As developers and users, for we are all engaged in research activities ourselves, we are exposed directly to the consumers of our software: mostly Ph.D. and undergraduate students, technicians and postdocs, as well as industry technicians with highly speci c needs. In numerous occasions, we have found the consumers of image processing technologies completely misguided as to 1) the nature of the task, 2)
the reality of a digital image, 3) their reasons for performing a certain set of operations on images, and 4) the consequences in data alteration resulting from the operations. In addition, today's very large image data sets require customized software and batch processing strategies. In our experience, users with very special needs cannot be e ectively trained, much less the knowledge be transferred between institutions and across generations of researchers, without a corpus of prior knowledge on digital imaging. Providing such corpus lays within our reach and well within our interests.

imageCosts

No registration fee. Housing and travel must be taken care of by the student. Financial aid for EU students is available (masters and PhD).

Stipend possibilities

Financial aid for EU students is available (masters and PhD). Interested applicants must submit a budget with their planned expenses; there is no guarantee that all expenses will be covered, but students will be notified in advance.

Accommodation

Same hotel as congress or open to students' choice. Pilsen Parkhotel first class category available at CZK 1000 (EUR 40)/person and night, in double and triple rooms: includes breakfast, VAT, city fees, parking place, and fitness.

Registration

Registration is open until July 31st. Quorum is limited to a maximum of 30 students.

Lecture format

There will be one instructor for every 3 students. There will be brief (15 min) general opening notes on the topics to learn about in each session, and then 1.5 hours of hands-on with an instructor. The school will proceed in English. Example data will be provided. We encourage applicants to bring extensive samples of their own research data to analyze.

Location

Pilsen, Czech Republic. The exact venue will likely be within the campus of the University of Western Bohemia. Details to come.

Time schedule

September 9 to 12 (Wednesday to Saturday), 2009.

Sessions start at 9:30, and proceed in intervals of 1.5 hours.

9:00 - 10:30

11:00 - 12:00

13:30 - 16:00

The rest of the afternoon is open for specialized sessions agreed upon by instructors and students to cover special needs, or free.

Organizers and sponsors

Albert Cardona, Institute of Neuroinformatics, University of Zurich and ETH Zurich.

The Swiss INCF node

The German INCF node

The UK INCF node

image


 

 

This course is taking place independently from the Congress.