alignment of 3d articulate shapes

38
ALIGNMENT OF 3D ARTICULATE SHAPES

Upload: chaim

Post on 24-Feb-2016

53 views

Category:

Documents


0 download

DESCRIPTION

Alignment of 3D Articulate Shapes. Articulated registration. Input: Two or more 3d point clouds (possibly with connectivity information) of an articulated object, possibly under different poses. Goal: Appropriately align all these point clouds together. Papers. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Alignment of 3D Articulate Shapes

ALIGNMENT OF 3D ARTICULATE SHAPES

Page 2: Alignment of 3D Articulate Shapes

Articulated registration• Input: Two or more 3d point clouds (possibly with

connectivity information) of an articulated object, possibly under different poses.

• Goal: Appropriately align all these point clouds together

Page 3: Alignment of 3D Articulate Shapes

Papers• Automatic Registration for Articulated Shapes

• Global Registration of Dynamic Range Scans for Articulated Model Reconstruction

• KinectAvatar: Fully Automatic Body Capture Using a Single Kinect

Page 4: Alignment of 3D Articulate Shapes

Automatic Registration for Articulated Shapes

• Pipeline

Page 5: Alignment of 3D Articulate Shapes

Feature• Pre-compute per-vertex coordinate frames and feature

descriptors (spin image)• Coordinate Frames (local coordinate system)

Page 6: Alignment of 3D Articulate Shapes

Spin Image• For a pair of points <p, x> => <>

• Radially Symmetric

Page 7: Alignment of 3D Articulate Shapes

Spin Image• For a point p and a set of points

Page 8: Alignment of 3D Articulate Shapes

Spin Image• “Spin” – Radial Symmetry

Page 9: Alignment of 3D Articulate Shapes

Spin Image for articulated shape• Localized to small neighborhoods

Page 10: Alignment of 3D Articulate Shapes

Feature Matching: match spin images• mu is the median of the upper half of the measurements,

ml is the median of the lower half of the measurements, fs = mu − ml

Page 11: Alignment of 3D Articulate Shapes

Motions sampling• for each correspondence candidate ( p,u )

Where R and t are from the computed coordinate frames

Page 12: Alignment of 3D Articulate Shapes

Motion clustering• Mode finding on Lie Groups: mean shift

Mean shift [1,2]• Kernel Density Estimation• Move along the gradient of

the kernel function

[1] Y. Cheng, “Mean shift, mode seeking, and clustering,” IEEE TPAMI., vol. 17, no. 8, pp. 790 –799, aug 1995.[2] D. Comaniciu and P. Meer, “Mean shift: a robust approach toward feature space analysis,” IEEE TPAMI ,vol. 24, no. 5, pp. 603 –619, may 2002

Page 13: Alignment of 3D Articulate Shapes

Mean shift on Lie Groups• Six dimensional space:

• The twist coordinate

• Distance metric (used in the profile of mean shift kernel)• Original:

• Simplified:

O. Tuzel, R. Subbarao, and P. Meer, “Simultaneous multiple 3d motion estimation via mode finding on lie groups,” in Int. Conf. on Comput. Vision, vol. 1, oct. 2005, pp. 18 – 25 Vol. 1. 6

Page 14: Alignment of 3D Articulate Shapes

Candidate Motions

Page 15: Alignment of 3D Articulate Shapes

Assign candidate motions to all points• Treat each candidate motion as a label• Cast into a labeling problem, solved by graph cut:

• argmin{ dataCost + smoothnessConstraint }

• Data cost:

Page 16: Alignment of 3D Articulate Shapes

Assign candidate motions to all points• Treat each candidate motion as a label• Cast into a labeling problem, solved by graph cut:

• argmin{ dataCost + smoothnessConstraint }

• Smoothness constraint: preserve edge length

Page 17: Alignment of 3D Articulate Shapes

Assign candidate motions to all points• Treat each candidate motion as a label• Cast into a labeling problem, solved by graph cut:

• argmin{ dataCost + smoothnessConstraint }

• Symmetric smoothness constraint

Page 18: Alignment of 3D Articulate Shapes

Assign candidate motions to all points• Treat each candidate motion as a label• Cast into a labeling problem, solved by graph cut:

• argmin{ dataCost + smoothnessConstraint }

Why symmetric?

The non-symmetric solutionprefers to preserve the edge lengths

Page 19: Alignment of 3D Articulate Shapes

Results• See the paper

• Note: provides segmentation simultaneously

Page 20: Alignment of 3D Articulate Shapes

Global Registration of Dynamic Range Scans for Articulated Model Reconstruction

• Goal

Page 21: Alignment of 3D Articulate Shapes

Basic Idea• Given a set of range scans• Perform Pairwise Registration as initialization• Maintain a DSG (Dynamic Sample Graph, the final output)

• Initially being the first frame (reference frame)• For each coming frame F, perform global registration:

• Iterate:• Estimate transformations for each part of DSG to align with all frames

up to F• Update segmentation of the DSG with the transformations

• Update DSG with the new frame F

Page 22: Alignment of 3D Articulate Shapes

Algorithm

Page 23: Alignment of 3D Articulate Shapes

Algorithm

Previous paper

Page 24: Alignment of 3D Articulate Shapes

Algorithm

For all frames, “uniformly” sample the input point set as candidates [1]

[1] MITCHELL, D. P. 1991. Spectrally optimal sampling for distribution ray tracing. ACM SIGGRAPH.

Page 25: Alignment of 3D Articulate Shapes

Algorithm

Requires registration of DSG with the new frame

Page 26: Alignment of 3D Articulate Shapes

Register DSG with new frame• For each rigid part, uniformly blend all transformations of

the points in this part.

Page 27: Alignment of 3D Articulate Shapes

Algorithm

The number of points for this part is less than 5 or 10% of the part.

Page 28: Alignment of 3D Articulate Shapes

Algorithm

Page 29: Alignment of 3D Articulate Shapes

Global Registration• Iterate:

• Estimate transformations for each part of DSG to align with all frames up so far

• Update segmentation of the DSG with the transformations• Update DSG with the new frame F

Page 30: Alignment of 3D Articulate Shapes

Global registration• Part one: optimize transformations

ICP-likeScheme: iterative

Special part

Page 31: Alignment of 3D Articulate Shapes

Fitting cost

Page 32: Alignment of 3D Articulate Shapes

Joint Costs

Page 33: Alignment of 3D Articulate Shapes

Joints Estimation• On reference frame only• For all edges in DSG with end points crossing two rigid

parts (with transformations Ta and Tb):• Average all end points to get initial estimation • Refine initial estimation by least square fitting:

Page 34: Alignment of 3D Articulate Shapes

Global registration• Part two: optimize weights

• Similar to first paper: cast as labeling problem and use graph cut for optimization

• Discard small regions

Global registration done

Page 35: Alignment of 3D Articulate Shapes

Algorithm

Page 36: Alignment of 3D Articulate Shapes

Update DSG• Each time, start from an empty DSG• For each frame F

• Include points from the candidates set of F if• Not overlapped based on distance and normal• Has a valid weight via interpolation

• Weight interpolation (assign each point to a rigid part)• Why:

• To add new points from the new frame• To reflect the update in transformations and label during global

registration• How:

• Compute distance to previous DSG, used as scores• Use the part form highest score, if it is significant compared to other

scores

Page 37: Alignment of 3D Articulate Shapes

Algorithm

MANSON, J., PETROVA, G., AND SCHAEFER, S. 2008. Streaming surface reconstruction using wavelets. Comput. Graph. Forum (Proceedings of SGP) 27, 5, 1411–1420.

Page 38: Alignment of 3D Articulate Shapes

Results• See Paper