discovering panoramas in web videos

24
Discovering Panoramas in Web Videos ACM Multimedia 2008 Feng Liu 1 , Yuhen-Hu 1,2 and Michael Gleicher 1

Upload: cecil

Post on 24-Feb-2016

71 views

Category:

Documents


0 download

DESCRIPTION

Discovering Panoramas in Web Videos. ACM Multimedia 2008 Feng Liu 1 , Yuhen-Hu 1,2 and Michael Gleicher 1. Outline. Introduction Video analysis Discovering panoramas Panorama synthesis Experiments Conclusion. Creating panoramic imagery. STEP 1 image alignment STEP2 image stitching. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Discovering Panoramas in Web Videos

Discovering Panoramas in Web Videos

ACM Multimedia 2008Feng Liu1, Yuhen-Hu1,2 and Michael Gleicher1

Page 2: Discovering Panoramas in Web Videos

Introduction Video analysis Discovering panoramas Panorama synthesis Experiments Conclusion

Outline

Page 3: Discovering Panoramas in Web Videos

STEP 1 image alignment

STEP2 image stitching

Creating panoramic imagery

Page 4: Discovering Panoramas in Web Videos

Not all video has appropriate sources◦ Not cover a wide field-of-view of a scene◦ Motion may be randomly◦ Image quality

Image sources in a video

Page 5: Discovering Panoramas in Web Videos

Three parts◦ video analysis◦ panorama source selection◦ panorama synthesis

Purpose-Discover Panoramas in Video

Page 6: Discovering Panoramas in Web Videos

Background-image homography

transformation

Page 7: Discovering Panoramas in Web Videos

Feature matching – SIFT Compute homography parameters – RANSAC

algo◦ Run k times:

(1)draw n samples randomly (2) fit parameters Θ with these n samples (3) for each of other N-n points, calculate its

distance to the fitted model, count the number

of inlier points, c◦ Output Θ with the largest c

Background-compute homography

Page 8: Discovering Panoramas in Web Videos

Example:line fittingn=2

c=3 c=15

…………………

Page 9: Discovering Panoramas in Web Videos

Image homography◦ Points should match◦ Measure error distance and give penalty

Moving object detect◦ For activity synopsis◦ examining the discrepancy between its local

motion vector and the global motion

Video analysis (1)

Page 10: Discovering Panoramas in Web Videos

Visual quality measures

Video analysis(2)

Method of [31]Tong et al 04

Method of [35]Wang et al 02

Average differencesacross block boundaries.

Page 11: Discovering Panoramas in Web Videos

Good panoramas◦ Good homography between frames◦ Video have high image quality◦ Cover a wild field view

Collision◦ More frame more wild field of view◦ More frame more accumulate error to degrade

quality

Discovering panoramas(1)

, vistual quality, extent of the scene

Page 12: Discovering Panoramas in Web Videos

Visual quality measure

Discovering panoramas(2)

Page 13: Discovering Panoramas in Web Videos

Scene extent measure

Discovering panoramas(3)

Reference Reference

Page 14: Discovering Panoramas in Web Videos

An Approximate Solution Steps◦ 1.Fetch a segment Sk from pool Sp◦ 2.Find the scene extent of Sk and corresponding

reference frame.◦ 3.Append the panorama set according to

equation(2). until .

◦ 4.If the scene meet , , add remainder to pool Sp.

◦ 5.If pool Sp != Ο , go to loop 1

Discovering panoramas(4)

Page 15: Discovering Panoramas in Web Videos

Discovering panoramas(4)shot boundary

segments

video

divide segments that have too penalty

Repeat until done

Discard those extent with too little coverage <

Page 16: Discovering Panoramas in Web Videos

Scene panorama synthesis◦ blending – feathering

◦ median-bilateral filtering

Panorama synthesis(1)

Page 17: Discovering Panoramas in Web Videos

Panorama synthesis(2) Activity synopsis synthesis

Detect

Discard

Select and composite into scene

Page 18: Discovering Panoramas in Web Videos

YouTube Travel and Events category – West Lake http://www.youtube.com/watch?v=6FKCHLfTns8&feature=player_embedded#!

◦ size 320 x 240

Experiments (1)

Page 19: Discovering Panoramas in Web Videos

Query panorama from YouTube 6 query , top 10 videos 86.7% contain panoramas

Experiments (2)

Page 20: Discovering Panoramas in Web Videos

Example

Page 21: Discovering Panoramas in Web Videos

Example

Page 22: Discovering Panoramas in Web Videos

ExampleNotre Dame, Paris

Page 23: Discovering Panoramas in Web Videos

In this paper, we presented an automatic method to discover panorama sources from casual videos.

“Query panoramas from YouTube”supports our proposal of using web videos as panorama source.

More importantly, this method contribute to presenting or summarizing imagery databases using panoramic imageries by mining the possible sources to synthesize the representations.

Conclusion

Page 24: Discovering Panoramas in Web Videos

[31] H. Tong, M. Li, H. Zhang, and C. Zhang. Blur detection for digital images using wavelet transform.In IEEE ICME, 2004.

[35] Z. Wang, G. Wu, H. Sheikh, E. Simoncelli, E.-H.Yang, and A. Bovik. Quality-aware images. IEEE Transactions on Image Processing, 15(6):1680 -1689,2006.

Original Videos: http://pages.cs.wisc.edu/~fliu/project/discover-pano.htm

Reference