CK1 Intelligent Surface Modeler

1
CK1Intelligent Surface Modeler

• By
• Yu Wing TAI
• Kam Lun TANG
• Advised by
• Prof. Chi Keung TANG

2
Overview of presentation

• Motivation
• Tensor Voting Algorithm
• Implementation
• Results
• Conclusion

• Motivation
• Tensor Voting Algorithm
• Implementation
• Results
• Conclusion

• theory and applications

• tensor and voting

• data representation and communication

• bunny
• dragon
• etc

Presented by Yu Wing TAI Kam Lun TANG
3
Motivation

• Theoretical interest
• Emulate human visual perception
• Applications
• 3D modeling

Presented by Yu Wing TAI Kam Lun TANG
4
Overview of presentation

• Motivation
• Tensor Voting Algorithm
• Implementation
• Results
• Conclusion

Presented by Yu Wing TAI Kam Lun TANG
5
What is Tensor Voting?

• Representation
• Constraint propagation
• Data communication

TENSOR
VOTING FIELDS
VOTING ALGORITHM
Presented by Yu Wing TAI Kam Lun TANG
6
Tensor Ellipse
SMOOTH CURVE
POINT JUNCTION

ELLIPSE (TENSOR)
Presented by Yu Wing TAI Kam Lun TANG
7
Tensor Ellipse
-

• Ball Tensor - 100 uncertainty in all directions

• Stick Tensor - 100 certainty in normal
  directions

Presented by Yu Wing TAI Kam Lun TANG
8
2D Stick Voting Field

• Encode smoothness

?
Presented by Yu Wing TAI Kam Lun TANG
9
2D Ball Voting Field

• Derived from 2D stick voting field
• Rotation and integration

Presented by Yu Wing TAI Kam Lun TANG
10
Voting Algorithm
Each input site propagates its information in a
neighborhood
voting summation of tensor votes accumulated in
a neighborhood
Presented by Yu Wing TAI Kam Lun TANG
11
3D Tensor Voting
Presented by Yu Wing TAI Kam Lun TANG
12
Results

• Noisy data
• Sparse data
• Large scale reconstruction
• Efficient neighborhood searching in 3D space
• Code Optimization
• Qualitative and quantitative analysis

• Noisy data
• Sparse data
• Large scale reconstruction
• Efficient neighborhood searching in 3D space
• Code Optimization
• Qualitative and quantitative analysis

Presented by Yu Wing TAI Kam Lun TANG
13
Result Robustness
14
Large scale reconstruction
1,153,856 triangles
35974 points
Presented by Yu Wing TAI Kam Lun TANG
15
Conclusion

• Intelligent surface modeler
• 3D surface description
• Tensor voting
• Results
• Robustness
• Large scale reconstruction
• Future work
• Multiscale feature segmentation and extraction

Presented by Yu Wing TAI Kam Lun TANG
16

• Thank you
• QA