1. The Brain from retina to extrastriate cortex

2. Neural processing responsible for vision photoreceptors retina –bipolar and horizontal cells –ganglion cells (optic nerve) optic nerves optic chiasma (X) lateral geniculate body striate cortex extrastriate cortex

3. Lateral inhibition Edge detection and contrast enhancement Bipolar, Horizontal and Ganglion cells

4. Lateral inhibition If no activity in neighboring photoreceptors, output = output of photoreceptor

5. Lateral inhibition if activity in neighboring photoreceptors, –output is decreased, possibly absent

6. Lateral inhibition via addition and negative weights

7. Optic nerve axons of the ganglion cells –1 million optic nerves –120 million photoreceptors

8. brains From http://www.marymt.edu/~psychol/cortex.html

9. From light to vision Lateral Geniculate Nucleus (LGN) Striate Cortex Geniculo-Striate Pathway

10. (LGN) Striate Cortex Striate cortex (primary visual centre) Neurons are edge detectors fires when an edge of a particular orientation is present

11. (LGN) Striate Cortex Striate cortex (primary visual centre) Neurons are edge detectors fires when an edge of a particular orientation is present frequent output vertical bar

12. (LGN) Striate Cortex Striate cortex (primary visual centre) Neurons are edge detectors fires when an edge of a particular orientation is present infrequent output diagonal bar

13. Edge detection each cell “tuned” to particular orientation –vertical –horizontal –diagonal cats: only horizontal and vertical humans: 10 degree steps edges at particular orientations and positions

14. Extrastriate cortex (beyond the striate cortex) V1

15. Extrastriate cortex Each area handles separate aspect of visual analysis –“V1-V2 complex”: Map for edges –V3: Map for form and local movement –V4: Map for colour –V5: Map for global motion Each is a visual module –connects to other areas –operates largely independently

16. Douglas A. Lyon, Ph.D. Chair, Computer Engineering Dept. Fairfield University, CT, USA Lyon@DocJava.com, http://www.DocJava.comhttp://www.DocJava.com Copyright 2002 © DocJava, Inc.

17. Background It is easy to find a bad edge! We know a good edge when we see it!

18. The Problem Given an expert + an image. The expert places markers on a good edge. Find a way to connect the markers.

19. Motivation Experts find/define good edges Knowledge is hard to encode.

20. Approach Mark an important edge Pixels=graph nodes Search in pixel space using a heuristic A* is faster than DP

21. Assumptions User is a domain expert Knowledge rep=heuristics+markers

22. Applications Photo interpretation Path planning (source+destination) Medical imaging

23. Photo Interpretation

24. Echocardiogram 3D-multi-scale analysis

25. Path Plans, the idea

26. Path Planning-pre proc. hist+thresh Dil+Skel

27. Path Planning - Search

28. The Idea The mind selects from filter banks The mind tunes the filters

29. Gabor filter w/threshold The Strong edge is the wrong edge!

30. Canny Edge Detector

31. Mehrotra and Zhang

34. Sub bands for 3x3 Robinson

35. Sub Bands 7x7 Gabor

36. Gabor-biologically motivated

37. Log filters=prefilter+laplacian

38. Mexican Hat (LoG Kernel)

39. The Log kernel

40. Oriented Filters are steerable

41. Changing Scale at 0 Degrees

42. Changing Phase at 0 degrees

43. Summary Heuristics+markers= knowledge Low-level image processing still needed Global optimization is not appropriate for all applications Sometimes we only want a single, good edge