Comparison with related works
on Flash/Long-exposure Image Integration

Back to the main page

This page shows various situations result

  1. Non-rigid motion
  2. Camera shake on inanimate object
  3. Illumination change & displacement
  4. Multiple Light Sources

“No-result” in the following images indicates the failure of correspondence search:
Conventional methods 2,3 use a correspondence algorithm of 2. However, the correspondence algorithm cannot get correspondences in some image pairs, and therefore the methods do not have a result.

Click an image to enlarge:
The following thumbnails are JPEG images. By clicking them, PNG images are shown.

I. Non-rigid motion


  (a) Flash image (b) Long-exposure
image
(c) Pitié et al. 1 (d) HaCohen et al. 2 (e) Hwang et al. 3 (f) Our method
I-1
I-2
I-3
I-4
I-5
I-6
I-7
I-8
I-9

Back to top

II. Camera shake on inanimate object


  (a) Flash image (b) Long-exposure
image
(c) Pitié et al. 1 (d) HaCohen et al. 2 (e) Hwang et al. 3 (f) Our method
II-1
II-2
II-3
II-4
II-5
II-6

Back to top

III. Illumination change & displacement


  (a) Flash image (b) Long-exposure
image
(c) Pitié et al. 1 (d) HaCohen et al. 2 (e) Hwang et al. 3 (f) Our method
III-1
III-2

Back to top

IV. Multiple light Sources


  (a) Flash image (b) Long-exposure
image
(c) Pitié et al. 1 (d) HaCohen et al. 2 (e) Hwang et al. 3 (f) Our method
IV-1
IV-2
IV-3

Back to top

Go to the other pages

References

  1. F. Pitié, A. Kokaram, and R. Dahyot, “Automated colour grading using colour distribution transfer,”
    Elsevier J. Comput. Vision Image Understanding, vol.107, no.1-2, pp.123-137, 2007.
  2. Y. HaCohen, E. Shechtman, D. B. Goldman, and D. Lischinski, “Non-rigid dense correspondence with applications for image enhancement,”
    ACM TOG (Proc. SIGGRAPH), vol.30, no.4, pp.70:1- 70:10, 2011.
  3. Y. Hwang, J. Y. Lee, I.S. Kweon, and S. J. Kim, “Color transfer using probabilistic moving least squares,”
    in Proc. IEEE CVPR, pp.3342-3349, June 2014.