1 / 36

CSE 554 Lecture 2: Skeleton and Thinning (Part I)

CSE 554 Lecture 2: Skeleton and Thinning (Part I). Fall 2013. Review. Binary pictures Created from grayscale images Basic operations Connected component labeling Morphological operators. Shape analysis. Questions about shapes: Metrics: length? Width? orientation ? …

tam
Download Presentation

CSE 554 Lecture 2: Skeleton and Thinning (Part I)

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. CSE 554Lecture 2: Skeleton and Thinning(Part I) Fall 2013

  2. Review • Binary pictures • Created from grayscale images • Basic operations • Connected component labeling • Morphological operators

  3. Shape analysis • Questions about shapes: • Metrics: length? Width? orientation? … • App: bio-shape analysis • What are the parts? • How similar are two shapes? • App: optical character recognition (OCR); model retrieval Human colon Microfilaments in a cell T H E OCR Model retrieval from a database

  4. Skeletons • Geometry at the center of the object • Compact, and capturing protruding shape parts Skeleton of 3D shapes: 1D curves and 2D surfaces Skeleton of 2D shapes: 1D curves

  5. Applications • Computer graphics and vision • Optical character recognition (a) • Shape comparison and retrieval (b) • Animation control (c) • … • Bio-medical image analysis • Vessel network analysis (d) • Virtual colonoscopy (e) • Protein modeling (f) • … (a) (b) (c) (d) (e) (f)

  6. Medial Axes (MA) • Interior points with multiple closest points on the boundary • So that it is “centered” in the object. Object MA

  7. Medial Axes (MA) • Properties  Thin • MA are curves (1D) in a 2D object, and surfaces (2D) in a 3D object. 3D MA 2D MA

  8. Medial Axes (MA) • Properties  Preserves object’s shape • The object can be reconstructed from MA and its distances to the boundary

  9. Medial Axes (MA) • Properties  Preserves object’s topology • 2D: # of connected components of object and background • 3D: # of connected components of object and background, and # of tunnels A 2D shape with 1 object component and 2 background components A 3D shape with 5 tunnels

  10. Medial Axes (MA) • Properties Not stable under boundary perturbation

  11. Skeletons • Approximation of medial axes • Retains significant parts of the medial axes • Varied definition for applications • In 3D, some apps prefer curve skeletons, and others prefer surface skeletons Animation: curve skeletons Virtual colonoscopy: curve skeletons Shape retrieval: surface skeletons Protein modeling: surface skeletons

  12. Computing Skeletons • Many methods for creating skeletons • A classical method on binary pictures: thinning • Can create curve skeletons in 2D and curve or surface skeletons in 3D • Relatively simple to implement • What we will cover: • Thinning on binary pictures (this lecture) • Simple in 2D, harder in 3D • Thinning on cell complexes (next lecture) • Simple in both 2D and 3D

  13. Medial Axes (MA) • Grassfire analogy: • Let the object represent a field of grass. A fire starts at the field boundary, and burns across the field at uniform speed. • MA are where the fire fronts meet. Object MA

  14. Medial Axes (MA) • Grassfire analogy: • Let the object represent a field of grass. A fire starts at the field boundary, and burns across the field at uniform speed. • MA are where the fire fronts meet.

  15. 2D Thinning • Iterative process that reduces a binary picture to a skeleton • Simulating the “grassfire burning” that defines MA Thinning on a binary picture

  16. 2D Thinning • Iterative process that reduces a binary picture to a skeleton • Simulating the “grassfire burning” that defines MA Thinning on a binary picture

  17. 2D Thinning • Thinning vs. morphological erosion • Iterative erosion eventually eliminates the object, but thinning preserves key pixels (voxels) so that the shape and topology of the object is retained Thinning Iterative erosion

  18. 2D Thinning • Curve-endpixels • Object pixels lying at the ends of curves, whose removal would shrink the skeleton (and hence losing shape information). c Curve-end pixel

  19. 2D Thinning • Curve-end pixels criteria • Object pixel c is a curve-end pixel if and only if c has exactly one connected pixel in the object. c Curve-end pixel and its connected pixel

  20. 2D Thinning • Simplepixels • Object pixels whose removal from the object does not change topology (i.e., # of components of object and background) Simple! 2 1 4 1 2 3 O:1 B:1 O:2 B:1 4 O:1 B:1 (using 8-connectivity for object) 3 O:1 B:2 O:1 B:1

  21. 1 1 2 2 4 4 3 3 2D Thinning • Simple pixels criteria • Object pixel p is simple if and only if setting p to background does not change the number of connected components of either the object or background in the 3x3 neighborhood of p. Simple! 4 O:1 B:1 O:1 B:1 O:1 B:2 O:2 B:1 1 2 3 (using 8-connectivity for object) O:1 B:2 O:1 B:3 O:1 B:1 O:1 B:1

  22. 2D Thinning • Simple pixels criteria • Object pixel p is simple if and only if setting p to background does not change the number of connected components of either the object or background in the 3x3 neighborhood of p. s s s s s s s s All simple pixels

  23. 2D Thinning • Border pixels • Pixels lying on the border of the object • To be considered for removal at each thinning iteration

  24. 2D Thinning • Border pixels criteria • Object pixel p is on the border if and only if p is connected to some background pixel b A border pixel and its 8-connected background pixels

  25. 2D Thinning • Border pixels criteria • Object pixel p is on the border if and only if p is connected to some background pixel b b b b b b b b b b Border pixels for 8-connectivity

  26. s s s b b b s c b b b b s b b b s s s 2D Thinning • Putting together: Removable pixels • Border pixels that are simple and not curve-end Border pixels Simple pixels Curve-end pixels Removal pixels

  27. 2D Thinning • Algorithm (attempt) 1 • Simultaneous removal of all removable points (“Parallel thinning”) • // Parallel thinning on a binary image I • Repeat: • Find all removable object pixels S in I • If S is empty, Break. • Set all pixels in S to be background in I • Output I

  28. 2D Thinning • Algorithm (attempt) 1 • Simultaneous removal of all removable points (“Parallel thinning”) • Does not work: the object topology is changed. How does this happen? Naïve parallel thinning

  29. 2D Thinning • A closer look at simple pixels • No longer “simple” when removed together • Naïve parallel thinning is not topology-preserving; more sophisticated strategies are needed • see Further Readings slide s s s s s s s s A simple example of how removing multiple simple pixels changes the object connectivity

  30. 2D Thinning • Algorithm 2 • Sequentially visit each removable pixel and check its simple-ness before removing the pixel. (“Serial Thinning”) • // Serial thinning on a binary image I • Repeat: • Find all removable object pixels S in I • If S is empty, Break. • Repeat for each pixel x in S in some order: • If x is simple, set x to be background in I • Output I

  31. 2D Thinning • Algorithm 2 • Sequentially visit each removable pixel and check its simple-ness before removing the pixel. (“Serial Thinning”) Serial thinning

  32. 2D Thinning • Algorithm 2 • Sequentially visit each removable pixel and check its simplicity before removing the pixel. (“Serial Thinning”) • Result is affected by the visiting “sequence” Serial thinning with two different visiting sequences of removable pixels

  33. 3D Thinning • Identifying removable voxels • Border voxels • Similar to 2D: voxels lacking a complete connected neighborhood in object • Simple voxels • Harder to characterize than 2D: Maintaining # of connected components is not sufficient (need to consider # of tunnels too) • Curve-end and surface-end voxels • Much harder to describe than 2D • Often uses a table of templates x Setting voxel x to background creates a “tunnel” in the object

  34. 3D Thinning • Two kinds of skeletons • Curve skeletons: only curve-end voxels are preserved during thinning • Surface skeletons: both curve-end and surface-end voxels are preserved • Thinning strategies • Serial or parallel • See Further Readings Object Curve skeleton Surface skeleton Method of [Palagyi and Kuba, 1999]

  35. Skeleton Pruning • Thinning is sensitive to boundary noise • Due to the instability of medial axes • Skeleton pruning • During thinning • E.g., using more selective criteria for end pixels (voxels) • After thinning • E.g., based on branch length • See Further Readings Object with boundary noise Resulting skeleton

  36. Further Readings on: Binary Pictures, MA and Thinning • Books • “Digital Geometry: geometric methods for digital picture analysis”, by Klette and Rosenfeld (2004) • “Medial representations: mathematics, algorithms and applications”, by Siddiqi and Pizer (2008) • Papers • “Digital topology: introduction and survey”, by Kong and Rosenfeld (1989) • Theories of binary pictures • “Thinning methodologies - a comprehensive survey”, by Lam et al. (1992) • A survey of 2D methods • “A Parallel 3D 12-Subiteration Thinning Algorithm”, by Palagyi and Kuba (1999) • Includes a good survey of 3D thinning methods • “Pruning medial axes”, by Shaked and Bruckstein (1998) • A survey of MA and skeleton pruning methods

More Related