Presentation is loading. Please wait.

Presentation is loading. Please wait.

Yannick FranckenChris HermansPhilippe Bekaert Hasselt University – tUL – IBBT Expertise Centre for Digital Media, Belgium

Similar presentations


Presentation on theme: "Yannick FranckenChris HermansPhilippe Bekaert Hasselt University – tUL – IBBT Expertise Centre for Digital Media, Belgium"— Presentation transcript:

1 Yannick FranckenChris HermansPhilippe Bekaert Hasselt University – tUL – IBBT Expertise Centre for Digital Media, Belgium {firstname.lastname}@uhasselt.be

2 Goal Geometric calibration of a camera w.r.t. a screen

3 Vision based HCI 3D reconstruction Motivation [Chen et al., SPIE 2002][Gorodnichy et al., SPIE 2002] [Francken et al., CVPR 2008][Nehab et al., CVPR 2008]

4 Planar mirror Related Work [Funk and Yang, CRV 2007] [Bonfort et al., ACCV 2006]

5 Planar mirror Spherical mirror –Corner reflections Related Work [Tarini et al., Graphical Models 2005]

6 Planar mirror Spherical mirror –Corner reflections –Edge reflections Related Work [Francken et al., CRV 2007]

7 Planar mirror Spherical mirror –Corner reflections –Edge reflections –Surface reflection Increased accuracy Less manual interventions Robust screen reflection detection Our Approach

8 Concept 1.Mirror detection 2.Screen pixel labeling 3.3D reconstruction

9 Mirror detection 1.Internal camera parameters K 2.Background subtraction 3.Edge extraction 4.Ellipse fitting 5.2D ellipse to 3D sphere

10 Screen pixel labeling

11

12

13

14

15

16

17

18 Reflection mask

19

20

21

22

23

24

25

26 3D reconstruction Reflected ray intersections Plane estimation Grid estimation Known parameters:

27 Reflected ray intersections Plane estimation Grid estimation Result: 2D pixel u  3D location x x = M. u 3D reconstruction Solution: Find 2D – 2D similarity transform

28 Overview x = M. u

29 Error as function of pattern refinement Results Accuracy –Ground truth –[Francken et al., CRV 2007] –Our approach

30 Error as function of sphere combinations Results

31 Error as function of sphere combinations Results

32 Error as function of sphere combinations Results

33 Error as function of sphere combinations Results

34 Screen-camera calibration using Gray codes –Increased accuracy –Less manual interventions –Robust screen reflection detection Conclusion

35 Gradient patterns –Speed! –Quality? Camera defocus –Which patterns are robust? Future Work

36 Questions? yannick.francken@uhasselt.behttp://research.edm.uhasselt.be/~yfrancken x = M. u

37


Download ppt "Yannick FranckenChris HermansPhilippe Bekaert Hasselt University – tUL – IBBT Expertise Centre for Digital Media, Belgium"

Similar presentations


Ads by Google