PyConZA 2013: "A random walk in image processing with scikit-image" by Stefan van der Walt

SciPy and NumPy are at the core of the scientific Python toolstack. Built on top of these are several add-on packages called "scikits", of which scikit-image is one of the more popular.

In this talk we discuss the motivation and development methodology behind scikit-image, highlight potential applications, and give an interactive overview of its capabilities.

The talk is targeted at anyone who enjoys graphics, or who likes to see how scientific tools are applied to solve real-world problems.

Transcript

1. Stéfan van der Walt @stefanvdwalt A random walk in image processing with scikit-image Applied Maths, Stellenbosch U. 1 / 25
2. s c i k i t - i m a g e is a scientific toolkit meant for use inside the scientific Python (scipy) eco-system. It is built on top of standard numpy arrays, and extends the functionality available in s c i p y . n d i m a g e . It aims to provide a platform for: 1. Reproducible image processing research 2. Industry application and quick evaluation of new algorithms 3. Education http://scikit-image.org 2 / 25
3. Motion deblur 3 / 25
4. 4 / 25
5. Super-resolution input 5 / 25
6. Super-resolution output 6 / 25
7. 7 / 25
8. 8 / 25
9. 9 / 25
10. Goals - Reproducible (image processing) research - Industry application - Rapid evaluation of new algorithms - Education 10 / 25
11. Reproducible research: sharkfin analysis In s k i m a g e , design choices mean pipelines Just Work (TM). 11 / 25
12. Reproducible research: sharkfin analysis 12 / 25
13. Reproducible research: sharkfin analysis Shortest paths 13 / 25
14. Industry application: chromosome segmentation Credit: Juan Nunez-Iglesias 14 / 25
15. Industry application: chromosome segmentation Credit: Juan Nunez-Iglesias 15 / 25
16. Industry application: chromosome segmentation Credit: Juan Nunez-Iglesias 16 / 25
17. Quick evaluation 17 / 25
18. Education The source code is open, well documented, easy to understand--instructive. In addition, n o v i c e module used by SWC for teaching Python: > > > f o r p i x e l i n p i c t u r e : . . . i f ( p i x e l . r e d > 1 2 8 ) a n d ( p i x e l . x < p i c t u r e . w i d t h ) : . . . p i x e l . r e d / = 2 18 / 25
19. Google Summer of Code 2013 19 / 25
20. Google Summer of Code 2013 Google Summer of Code 2013 20 / 25
21. Further examples: sun warping 21 / 25
22. Further examples: LAB histogram 22 / 25
23. The Future from Meunier, Lambiotte and Bullmore 23 / 25
24. Super hexagon! Credit: Laurens Leerink 24 / 25
25. Please join us on the scikit-image mailing list! http://scikit-image.org The End 25 / 25