Separable Kernel for Image Deblurring
Abstract
The blur effects caused by camera shake and object motion in scenes occur frequently in photography, producing disappointing blurry images with inevitable information loss, which becomes one of the most common reasons for discarding photographs. In this talk, I would like to present some of our recent results on single image blind deblurring. We investigate it in a completely new perspective by proposing separable kernel to represent the inherent properties of the camera and scene system. Specifically, we decompose a blur kernel into three individual descriptors (trajectory, intensity and point spread function) so that they can be optimized separately. To demonstrate the advantages, we extract one-pixel-width trajectories of blur kernels and propose a random perturbation algorithm to optimize them but still keeping their continuity. For many cases, where current deblurring approaches fall into local minimum, excellent deblurred results and correct blur kernels can be obtained by individually optimizing the kernel trajectories. Our work strongly suggests that more constraints and priors should be introduced to blur kernels in solving the deblurring problem because blur kernels have lower dimensions than images.
SpeakerProf. Lu FANG | |
Date & Time5 Nov 2014 (Wednesday) 11:00 - 12:00 | |
VenueE11 - 4045 (University of Macau) | |
Organized byDepartment of Computer and Information Science |
Biography
Dr. Lu Fang is currently an Associate Professor in the Department of Electronic Engineering and Information Science, the University of Science and Technology of China (USTC). She used to be a Post-doc Research Fellow from 2011-2012 in Hong Kong University of Science and Technology (HKUST) and Singapore University of Technology and Design (SUTD) respectively. She was a Visiting Scholar at Northwestern University in 2010, and a Visiting Professor in Microsoft Research Asia (MSRA) in 2013.
Dr. Fang received her Ph.D. in Electronic and Computer Engineering from HKUST in 2011, and B.E. in Electronic Engineering and Information Science from the USTC in 2007, respectively. Dr. Fang’s research interests include multimedia processing, image/video coding, and computer vision. Dr. Fang has published over 50 IEEE international conference and IEEE journal publications, including IEEE SPM, IEEE TIP, IEEE TCSVT, IEEE TMM, IEEE CVPR etc. She used to be awarded for Humboldt Research Fellowship for Experienced Researchers, Best Paper Candidate in ICME 2011, HKTIIT Post-Graduate Excellence Scholarship in 2011. Dr. Fang has served as APSIPA Social Net Committee, Publication Chair of IEEE ICME 2017, Finance Chair of 18th International Packet Video Workshop (PV 2010). She also served as TPC members for several conferences such as IEEE ICIP 2014, IEEE ICME 2014, IEEE ICASSP 2014 et al.