From local to global: Edge profiles to camera motion in blurred images

Subeesh Vasu    A.N. Rajagopalan

Input blurred image	Latent image prediction	Edge profile image	Deblurred output

Abstract

In this work, we investigate the relation between the edge profiles present in a motion blurred image and the underlying camera motion responsible for causing the motion blur. While related works on camera motion estimation (CME) rely on the strong assumption of space-invariant blur, we handle the challenging case of general camera motion. We first show how edge profiles `alone' can be harnessed to perform direct CME from a single observation. While it is routine for conventional methods to jointly estimate the latent image too through alternating minimization, our above scheme is best-suited when such a pursuit is either impractical or inefficacious. For applications that actually favor an alternating minimization strategy, the edge profiles canserve as a valuable cue. We incorporate a suitably derived constraint from edge profiles into an existing blind deblurring framework and demonstrate improved restoration performance. Experiments reveal that this approach yields state-of-the-art results for the blind deblurring problem.

Technical Paper, Supplementary, Poster, and Bibtex

Experimental Results

Synthetic example

Blurred image	Xu et al. CVPR 2013	Ours - EpAlone	Ours - Xu + Ep
	Estimated kernels - Xu et al. CVPR 2013	Estimated kernels - EpAlone	Estimated kernels - Xu + Ep

Real examples

Blurred image	Whyte et al. CVPR 2010	Xu et al. CVPR 2013	Pan et al. CVPR 2016	Ours - EpAlone	Ours - Xu + Ep

Blurred image	Gupta et al. ECCV 2010	Xu et al. CVPR 2013	Ours - EpAlone	Ours - Xu + Ep

Blurred image	Gupta et al. ECCV 2010	Xu et al. CVPR 2013	Ours - EpAlone	Ours - Xu + Ep

Blurred image	Gupta et al. ECCV 2010	Whyte et al. CVPR 2010	Xu et al. CVPR 2013	Pan et al. CVPR 2016	Ours - EpAlone	Ours - Xu + Ep

Blurred image	Gupta et al. ECCV 2010	Xu et al. CVPR 2013	Ours - EpAlone	Ours - Xu + Ep

Blurred image	Harmeling et al. NIPS 2010	Hirsch et al. ICCV 2011	Xu et al. CVPR 2013	Schuler et al. PAMI 2016	Ours - EpAlone	Ours - Xu + Ep

Blurred image	Whyte et al. CVPR 2010	Hirsch et al. ICCV 2011	Xu et al. CVPR 2013	Schuler et al. PAMI 2016	Ours - EpAlone	Ours - Xu + Ep

Blurred image	Gupta et al. ECCV 2010	Whyte et al. CVPR 2010	Xu et al. CVPR 2013	Pan et al. CVPR 2016	Ours - EpAlone	Ours - Xu + Ep

Blurred image	Harmeling et al. NIPS 2010	Xu et al. CVPR 2013	Ours - EpAlone	Ours - Xu + Ep

References

[1] O. Whyte, J. Sivic, A. Zisserman, and J. Ponce. “Non-uniform deblurring for shaken images”, CVPR 2010.

[2] A. Gupta, N. Joshi, L. Zitnick, M. Cohen, and B. Curless. “Single image deblurring using motion density functions”, ECCV 2010.

[3] S. Harmeling, H. Michael, and B. Sch ̈olkopf. “Space-variant single-image blind deconvolution for removing camera shake”, NIPS 2010.

[4] M. Hirsch, C. J. Schuler, S. Harmeling, and B. Sch ̈olkopf. “Fast removal of non-uniform camera shake”, ICCV 2011.

[5] L. Xu, S. Zheng, and J. Jia. “Unnatural l₀ sparse representation for natural image deblurring”, CVPR 2013.

[6] J. Pan, D. Sun, H. Pfister, and M.-H. Yang. “Blind image deblurring using dark channel prior”, CVPR 2016.

[7] C. J. Schuler, M. Hirsch, S. Harmeling, and B. Sch ̈olkopf. “Learning to deblur.”, PAMI 2016.

[8] Köhler, R, et al. “Recording and playback of camera shake: Benchmarking blind deconvolution with a real-world database.”, ECCV 2012.

[9] Lai, Wei-Sheng, et al. “A comparative study for single image blind deblurring.”, CVPR 2016.