improved census transforms for resource-optimized stereo vision

36
Improved Census Transforms for Resource-Optimized Stereo Vision Wade S. Fife, Member, IEEE, James K. Archibald, Senior Member, IEEE IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, VOL. 23, NO. 1, JANUARY 2013

Upload: kare

Post on 23-Feb-2016

102 views

Category:

Documents


0 download

DESCRIPTION

Improved Census Transforms for Resource-Optimized Stereo Vision. Wade S. Fife, Member, IEEE, James K. Archibald, Senior Member, IEEE. IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, VOL. 23, NO. 1, JANUARY 2013. Outline. Introduction Related Work Proposed Algorithm - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

Improved Census Transforms for Resource-Optimized Stereo Vision

Wade S. Fife, Member, IEEE, James K. Archibald, Senior Member, IEEE

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, VOL. 23, NO. 1, JANUARY 2013

Page 2: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

2

Outline• Introduction• Related Work• Proposed Algorithm

• Sparse Census Transform• Generalized Census Transform• Hardware Implementation

• Experimental Results• Conclusion

Page 3: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

3

Introduction

Page 4: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

4

Introduction• The challenges:

• The enormous amount of computation required to identify the corresponding points in the images.

• It is critical to…

• maximize the accuracy and throughput of the stereo system • while minimizing the resource requirements

Page 5: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

5

Objective• Propose the sparse census transforms :

• Reduce the resource requirements of census-based systems• Maintain correlation accuracy

• Propose the generalized census transforms :

• A new class of census-like transforms • Increase the robustness and flexibility

Page 6: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

6

Related Work

Page 7: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

7

Related Work• Census Transform :

• Color• Gradient

Page 8: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

8

Related Work• After aggregation step:

Census on colors Census on gradients

Page 9: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

9

Related Work• Sparse census[6] :

• Half of the bits

X

[6] C. Zinner, M. Humenberger, K. Ambrosch, and W. Kubinger, “An optimized software-based implementation of a census-based stereo matching algorithm,” in Proc. 4th ISVC, 2008, pp. 216–227.

The computation costs for the hamming distances are quite large.

Page 10: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

10

Related Work• Mini-census[8] :

X

[8] N.-C. Chang, T.-H. Tsai, B.-H. Hsu, Y.-C. Chen, and T.-S. Chang,“Algorithm and architecture of disparity estimation with mini-census adaptive support weight,” IEEE Trans. Circuits Syst. Video Technol., vol. 20, no. 6, pp. 792–805, Jun. 2010.

Page 11: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

11

Related Work• Mini-census[8] :

• Mini-census adaptive support weight

[8] N.-C. Chang, T.-H. Tsai, B.-H. Hsu, Y.-C. Chen, and T.-S. Chang,“Algorithm and architecture of disparity estimation with mini-census adaptive support weight,” IEEE Trans. Circuits Syst. Video Technol., vol. 20, no. 6, pp. 792–805, Jun. 2010.

Page 12: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

12

Related Work• Mini-census[8] :

Page 13: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

13

ProposedAlgorithm

Page 14: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

14

Sparse Census Transform• Definition :

• N: the set of points within a T T window around p• : a new set of N•

P’

P

Page 15: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

15

Transform Point Selection• Goal : minimize the size of the census transform vector

• Challenge: Must quantify how much each point in the transform window contributes to overall correlation accuracy

• Test correlation accuracy:

• Define a sparse census transform consisting of a single point (| | = 1)• Determine how consistently this point leads to correct correlation• 13 13 correlation window (aggregation)

Page 16: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

16

Transform Point Selection• Go

Tsukuba Venus Average

Teddy Cones

Bright: Higher correlation accuracy

25 25 neighborhood

Page 17: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision
Page 18: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

18

Transform Point Selection• Further from the center : value decreasing

• Very near the center : less effective

• It is best to choose points that are neither too far from nor too close to the center pixel.

• Optimal distance : 2 pixels• If the image is noisy should be slightly further

from the center

Page 19: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

19

Transform Point Selection•

Tsukuba Venus Average

Teddy Cones

Bright: Higher correlation accuracy

37 37 neighborhood

Tsukuba Venus

Teddy Cones

With Gaussian noise( = 5.12)

Page 20: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

20

Proposed Sparse Census Transform• Very good correlation accuracy can be achieved using very sparse transforms.

16-point 12-point 8-point

4-point 2-point 1-point

Page 21: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

21

Experimental Results

Page 22: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

22

Generalized Census Transform• Goal : greater freedom in choosing the census transform design

• Definition : redrawing the transform as a graph

3 3 census

3 3 correlation(aggregation)

Page 23: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

23

Generalized Census Transform• As..

• (1)transform neighborhoods become more and more sparse• (2)fewer pixels are used in the correlation process

• selection of points to include in the transform becomes more critical

2-point 2-edge

Horizontal + Vertical

Page 24: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

24

Generalized Census Transform

symmetric

Page 25: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

25

Proposed Generalized Census Transform• Benefits :

• Often require a smaller census transform window (memory)• Increased robustness under varying conditions (noise)

16-edge 12-edge 8-edge

4-edge 2-edge 1-edge

Page 26: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

26

Experimental Results

Page 27: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

27

Experimental Results

Page 28: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

28

Hardware Implementation• Pipelining : to increase throughput in an FPGA implementation

(Field Programmable Gate Array)

Range : 0~3

3 2 1 0

3 2 1 0

3 2 1 0

3 2 1 0

3 2 1 0

One input pixel per clock cycle &Output one disparity result per clock cycle

Page 29: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

29

Hardware Implementation• Correlation window sum (Aggregation) :

Page 30: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

30

ExperimentalResults

Page 31: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

12-edge 4-edgeFull 7x7 censusGround TruthLeft Image

Page 32: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

12-edge 4-edgeFull 7x7 censusGround TruthLeft Image

Page 33: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

12-edge 4-edgeFull 7x7 censusLeft Image

Page 34: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

34

Experimental Results

LUTs (look-up tables) : the amount of logic required to implement the methodFFs : the number of 1-bit registers (the amount of pipelining used)RAMs : the number of 18-kbit block memoriesFreq. : the maximum operating frequency reported by synthesis

𝟖𝟖% ↓ 𝟔𝟏% ↓

Page 35: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

35

Conclusion

Page 36: Improved  Census Transforms  for  Resource-Optimized  Stereo Vision

36

Conclusion• Proposed and analyzed in this paper:

• A range of sparse census transforms

• reduce hardware resource requirements• attempting to maximize correlation accuracy.• often better than or nearly as good as the full census

• Generalized census transforms

• increased robustness in the presence of image noise