1. CSE 473/573 Computer Vision and Image Processing (CVIP) Ifeoma Nwogu inwogu@buffalo.edu Lecture 19 – Dense motion estimation (OF) 1

2. Schedule Last class – We finished stereo and multi-view geometry (high level) Today – Optical flow Readings for today: Forsyth and Ponce 10.6, 11.1.2 2

3. Visual motion Many slides adapted from S. Seitz, R. Szeliski, M. Pollefeys

4. Motion and perceptual organization Sometimes, motion is the only cue

5. Motion and perceptual organization Sometimes, motion is the only cue

6. Motion and perceptual organization Even “impoverished” motion data can evoke a strong percept G. Johansson, “Visual Perception of Biological Motion and a Model For Its Analysis", Perception and Psychophysics 14, 201-211, 1973.

7. Motion and perceptual organization Even “impoverished” motion data can evoke a strong percept G. Johansson, “Visual Perception of Biological Motion and a Model For Its Analysis", Perception and Psychophysics 14, 201-211, 1973.

8. Motion and perceptual organization Even “impoverished” motion data can evoke a strong percept G. Johansson, “Visual Perception of Biological Motion and a Model For Its Analysis", Perception and Psychophysics 14, 201-211, 1973. YouTube video

9. Uses of motion Estimating 3D structure Segmenting objects based on motion cues Learning and tracking dynamical models Recognizing events and activities

10. Classes of techniques for motion estimation Feature-based methods – Extract visual features (corners, textured areas) and track them – Sparse motion fields, but possibly robust tracking – Suitable especially when image motion is large (10s of pixels) Direct-methods – Directly recover image motion from spatio-temporal image brightness variations – Global motion parameters directly recovered without an intermediate feature motion calculation – Dense motion fields, but more sensitive to appearance variations – Suitable for video and when image motion is small (< 10 pixels) Szeliski

11. Motion field The motion field is the projection of the 3D scene motion into the image

12. Patch based image motion How do we determine correspondences? Assume all change between frames is due to motion: I J

13. Optical flow Definition: optical flow is the apparent motion of brightness patterns in the image Ideally, optical flow would be the same as the motion field Have to be careful: apparent motion can be caused by lighting changes without any actual motion – Think of a uniform rotating sphere under fixed lighting vs. a stationary sphere under moving illumination

14. Estimating optical flow Given two subsequent frames, estimate the apparent motion field u(x,y) and v(x,y) between them Key assumptions Brightness constancy: projection of the same point looks the same in every frame Small motion: points do not move very far Spatial coherence: points move like their neighbors I(x,y,t–1)I(x,y,t)I(x,y,t)

15. Brightness Constancy Equation: Linearizing the right side using Taylor expansion: The brightness constancy constraint I(x,y,t–1)I(x,y,t)I(x,y,t) Hence,

16. The brightness constancy constraint How many equations and unknowns per pixel? – One equation, two unknowns What does this constraint mean? The component of the flow perpendicular to the gradient (i.e., parallel to the edge) is unknown

17. The brightness constancy constraint How many equations and unknowns per pixel? – One equation, two unknowns What does this constraint mean? The component of the flow perpendicular to the gradient (i.e., parallel to the edge) is unknown edge (u,v)(u,v) (u’,v’) gradient (u+u’,v+v’) If (u, v) satisfies the equation, so does (u+u’, v+v’) if

18. The aperture problem Perceived motion

19. The aperture problem Actual motion

20. The barber pole illusion http://en.wikipedia.org/wiki/Barberpole_illusion

21. The barber pole illusion http://en.wikipedia.org/wiki/Barberpole_illusion

22. Solving the aperture problem How to get more equations for a pixel? Spatial coherence constraint: pretend the pixel’s neighbors have the same (u,v) – E.g., if we use a 5x5 window, that gives us 25 equations per pixel B. Lucas and T. Kanade. An iterative image registration technique with an application toAn iterative image registration technique with an application to stereo vision.stereo vision. In Proceedings of the International Joint Conference on Artificial Intelligence, pp. 674–679, 1981.

23. Solving the aperture problem Least squares problem: B. Lucas and T. Kanade. An iterative image registration technique with an application toAn iterative image registration technique with an application to stereo vision.stereo vision. In Proceedings of the International Joint Conference on Artificial Intelligence, pp. 674–679, 1981. When is this system solvable? What if the window contains just a single straight edge?

24. Conditions for solvability “Bad” case: single straight edge

25. Conditions for solvability “Good” case

26. Lucas-Kanade flow Linear least squares problem B. Lucas and T. Kanade. An iterative image registration technique with an application toAn iterative image registration technique with an application to stereo vision.stereo vision. In Proceedings of the International Joint Conference on Artificial Intelligence, pp. 674–679, 1981. The summations are over all pixels in the window Solution given by

27. Lucas-Kanade flow Recall the Harris corner detector: M = A T A is the second moment matrix We can figure out whether the system is solvable by looking at the eigenvalues of the second moment matrix The eigenvectors and eigenvalues of M relate to edge direction and magnitude The eigenvector associated with the larger eigenvalue points in the direction of fastest intensity change, and the other eigenvector is orthogonal to it

28. Visualization of second moment matrices

30. Interpreting the eigenvalues 1 2 “Corner” 1 and 2 are large, 1 ~ 2 1 and 2 are small “Edge” 1 >> 2 “Edge” 2 >> 1 “Flat” region Classification of image points using eigenvalues of the second moment matrix:

31. Visualization of second moment matrices

32. The Aperture Problem Let Algorithm: At each pixel compute by solving M is singular if all gradient vectors point in the same direction e.g., along an edge of course, trivially singular if the summation is over a single pixel or there is no texture i.e., only normal flow is available (aperture problem) Corners and textured areas are OK and Szeliski

33. Example * From Khurram Hassan-Shafique CAP5415 Computer Vision 2003

34. Uniform region – gradients have small magnitude – small  1, small 2 – system is ill-conditioned

35. SSD – uniform region

36. Edge – gradients have one dominant direction – large  1, small 2 – system is ill-conditioned

37. SSD Surface -- edge

38. High-texture or corner region – gradients have different directions, large magnitudes – large  1, large 2 – system is well-conditioned

39. SSD Surface – textured area or corner

40. Optical Flow Results * From Khurram Hassan-Shafique CAP5415 Computer Vision 2003

41. Errors in Lucas-Kanade Fails when intensity structure in window is poor The motion is large (larger than a pixel) – Iterative refinement – Coarse-to-fine estimation – Exhaustive neighborhood search (feature matching) A point does not move like its neighbors – Motion segmentation Brightness constancy does not hold – Exhaustive neighborhood search with normalized correlation

42. image I image J JwJw warp refine + Pyramid of image JPyramid of image I image I image J Coarse-to-Fine Estimation u=10 pixels u=5 pixels u=2.5 pixels u=1.25 pixels Szeliski

43. J JwJw I warp refine + J JwJw I warp refine + J pyramid construction J JwJw I warp refine + I pyramid construction Coarse-to-Fine Estimation Szeliski

44. Multi-resolution registration * From Khurram Hassan-Shafique CAP5415 Computer Vision 2003

45. Optical Flow Results * From Khurram Hassan-Shafique CAP5415 Computer Vision 2003

46. Optical Flow Results * From Khurram Hassan-Shafique CAP5415 Computer Vision 2003

47. State-of-the-art optical flow Start with something similar to Lucas-Kanade + gradient constancy + energy minimization with smoothing term + region matching + keypoint matching (long-range) Large displacement optical flowLarge displacement optical flow, Brox et al., CVPR 2009 Region-based+Pixel-based +Keypoint-based Source: J. Hays

48. Feature tracking So far, we have only considered optical flow estimation in a pair of images If we have more than two images, we can compute the optical flow from each frame to the next Given a point in the first image, we can in principle reconstruct its path by simply “following the arrows”

49. Ambiguity of optical flow – Need to find good features to track Large motions, changes in appearance, occlusions, disocclusions – Need mechanism for deleting, adding new features Drift – errors may accumulate over time – Need to know when to terminate a track Tracking challenges

50. Shi-Tomasi feature tracker Find good features using eigenvalues of second- moment matrix – Key idea: “good” features to track are the ones whose motion can be estimated reliably From frame to frame, track with Lucas-Kanade – This amounts to assuming a translation model for frame-to- frame feature movement Check consistency of tracks by affine registration to the first observed instance of the feature – Affine model is more accurate for larger displacements – Comparing to the first frame helps to minimize drift J. Shi and C. Tomasi. Good Features to Track. CVPR 1994.Good Features to Track

51. Tracking example J. Shi and C. Tomasi. Good Features to Track. CVPR 1994.Good Features to Track

52. Non Gaussian noise Least square solution assumes error in the image motion estimation are Gaussian in nature The matrix M or the structured tensor matrix is computed using finite difference methods – forward, backward, and central differences – Can obtain higher order evaluations based on how the derivatives are computed (e.g adaptive windowing etc.) 52

53. Robust Estimation Noise distributions are often non-Gaussian, having much heavier tails. Noise samples from the tails are called outliers. Sources of outliers (multiple motions): – specularities / highlights – jpeg artifacts / interlacing / motion blur – multiple motions (occlusion boundaries, transparency) velocity space u1u1 u2u2 + + Black

54. Occlusion occlusiondisocclusionshear Multiple motions within a finite region. Black

55. Coherent Motion Possibly Gaussian. Black

56. Multiple Motions Definitely not Gaussian. Black

57. Layered Scene Representations

58. Motion representations How can we describe this scene? Szeliski

59. Block-based motion prediction Break image up into square blocks Estimate translation for each block Use this to predict next frame, code difference (MPEG-2) Szeliski

60. Layered motion Break image sequence up into “layers”:  = Describe each layer’s motion Szeliski

61. Layered motion Advantages: can better handle occlusions / disocclusions each layer’s motion can be smooth can be used for video segmentation in semantic processing Difficulties: how to determine the correct number of layers? how to assign pixels? how to model the layer motion? Szeliski

62. Layers for video summarization Szeliski

63. Background modeling (MPEG-4) Convert masked images into a background sprite for layered video coding + + + = Szeliski

64. What are layers? [Wang & Adelson, 1994; Darrell & Pentland 1991] intensities alphas velocities Szeliski

65. Fragmented Occlusion

66. Results

68. How to estimate the layers 1.compute coarse-to-fine flow 2.estimate affine motion in blocks (regression) 3.cluster with k-means 4.assign pixels to best fitting affine region 5.re-estimate affine motions in each region… Szeliski

69. Layer synthesis For each layer: stabilize the sequence with the affine motion compute median value at each pixel Determine occlusion relationships Szeliski

70. Results Szeliski

71. Recent GPU Implementation http://gpu4vision.icg.tugraz.at/ Real time flow exploiting robust norm + regularized mapping

72. Recent results: SIFT Flow

73. Slide Credits Svetlana Lazebnik – UIUC Trevor Derrell – UC Berkeley 73

74. Next class Segmentation via clustering Readings for next lecture: – Forsyth and Ponce chapter 9 – Szelinski chapter 5 Readings for today: – Forsyth and Ponce 10.6 and 11.1.2 74

75. Questions 75