3D content production St. Petersburg Electrotechnical University,  St. Petersburg , Russia Sergey A. Sukhanov
Outline <ul><li>What is 3D </li></ul><ul><li>3D  Technology </li></ul><ul><li>The need for 3D video content </li></ul><ul>...
What is 3D?
3D  Technology
3D Content Production Three main approaches to creating 3-D content: •  live camera capture •  computer generated imagery ...
Live camera capture Live camera capture:  recording stereo sequences of images   simultaneously A depth-range camera Speci...
D epth map Regular 2D color video Accompanying 8-bit depth-images with the  same spatio-temporal resolution The depth-imag...
2D to 3D conversion 2D to 3D conversion:  using monocular depth cues obtained from a 2D sequence to generate the equivalen...
D epth map s and  depth image based  r endering (DIBR)
D epth map s generation B lock-matching based motion estimation 4x4 block Using previous frame as reference
D epth map s generation <ul><li>identifies the chromatic homogeneous regions </li></ul><ul><li>group together pixels basin...
D epth map s generation Fusion <ul><li>- to enhance the block-based depth map </li></ul><ul><li>to eliminate the staircase...
D epth image based   rendering Camera configuration for generation of virtual stereoscopic images. 3D Image Warping <ul><l...
D epth image based   rendering Hole-Filling -  occlusion :  two different pixels of the real view image are warped to the ...
Conclusion R ising demand on new  3D content •  live camera capture •  2D to 3D conversion <ul><li>Advantages  of  propose...
<ul><li>Thank you for your attention! </li></ul>
Upcoming SlideShare
Loading in …5
×

Sergey A. Sukhanov, "3D content production"

1,506 views

Published on

Published in: Business, Technology
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
1,506
On SlideShare
0
From Embeds
0
Number of Embeds
223
Actions
Shares
0
Downloads
14
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Sergey A. Sukhanov, "3D content production"

  1. 1. 3D content production St. Petersburg Electrotechnical University, St. Petersburg , Russia Sergey A. Sukhanov
  2. 2. Outline <ul><li>What is 3D </li></ul><ul><li>3D Technology </li></ul><ul><li>The need for 3D video content </li></ul><ul><li>Live camera capture </li></ul><ul><li>2D to 3D conversion </li></ul><ul><li>Depth maps and depth image based rendering (DIBR) method </li></ul><ul><li>Conclusion </li></ul>
  3. 3. What is 3D?
  4. 4. 3D Technology
  5. 5. 3D Content Production Three main approaches to creating 3-D content: • live camera capture • computer generated imagery (CGI) • 2D to 3D conversion
  6. 6. Live camera capture Live camera capture: recording stereo sequences of images simultaneously A depth-range camera Special 3D cameras (single body) Special 3D cameras (single body) side-by-side rig side-by-side rig
  7. 7. D epth map Regular 2D color video Accompanying 8-bit depth-images with the same spatio-temporal resolution The depth-images are normalized to a near clipping plane Z near and a far clipping plane Z far .
  8. 8. 2D to 3D conversion 2D to 3D conversion: using monocular depth cues obtained from a 2D sequence to generate the equivalent 3D sequence 1. Cutting out pieces of the image and nudging them around 2 . Projecting an image on a 3D model 3 . Turning motion parallax into stereoscopy 4 . Using depth maps and DIBR process <ul><li>challenging task because of automatic segmentation </li></ul><ul><li>main purpose: to generate additional views </li></ul>
  9. 9. D epth map s and depth image based r endering (DIBR)
  10. 10. D epth map s generation B lock-matching based motion estimation 4x4 block Using previous frame as reference
  11. 11. D epth map s generation <ul><li>identifies the chromatic homogeneous regions </li></ul><ul><li>group together pixels basing on their likeness </li></ul>Color Segmentation
  12. 12. D epth map s generation Fusion <ul><li>- to enhance the block-based depth map </li></ul><ul><li>to eliminate the staircase effort </li></ul><ul><li>to assign better depth values in each region </li></ul>
  13. 13. D epth image based rendering Camera configuration for generation of virtual stereoscopic images. 3D Image Warping <ul><li>projects each pixel of the real view </li></ul><ul><li>into the 3D world </li></ul><ul><li>re-project these pixels back to 2D </li></ul><ul><li>image of the virtual view </li></ul>
  14. 14. D epth image based rendering Hole-Filling - occlusion : two different pixels of the real view image are warped to the same location in the virtual view - disocclusion : is due to the occluded area in the real view may become visible in the virtual view Left-eye view image created by 3D image warping with holes due to disocculsion Enlarged left-eye view image with stripe distortion after linear interpolation based hole-filling
  15. 15. Conclusion R ising demand on new 3D content • live camera capture • 2D to 3D conversion <ul><li>Advantages of proposed method: </li></ul><ul><li>Automation </li></ul><ul><li>use of a single view of the scene </li></ul><ul><li>effectiveness </li></ul><ul><li>Drawback s: </li></ul><ul><li>high computational requirement </li></ul><ul><li>- n ot suitable for real-time applications </li></ul>
  16. 16. <ul><li>Thank you for your attention! </li></ul>

×