Image Deblurring using Extra Image Pairs and Sharp Structure Priors

PhD Thesis Proposal Defence


Title: "Image Deblurring using Extra Image Pairs and Sharp Structure Priors"

by

Mr. Lu YUAN


Abstract:

The Recovery of a sharp version of the blurred image is a challenging problem 
in digital imaging. Previous works have achieved dramatic progress, yet the 
heavy ill-posedness of the problem leads to the results still far from perfect. 
In my thesis, I will explore latent and valuable information (or prior) from 
observations to provide a good condition for image deblurring.

The first idea comes from the help of additional correlative images. By 
combining information between blurred image and noisy image pair, we can 
estimate a very accurate blur kernel and restore a high-quality original image, 
which can not be obtained by simple single image denoising or single image 
deblurring. The idea further pushes me to develop a more general framework for 
image deblurring with a sequence of images, which do not limit to blurred/noisy 
image pair, and even can be multiple blurred image pairs. Our approach lies on 
the assumption that these blurred images with different blurs are derived from 
the same original image and different blurs will result in the loss of 
different frequency components during imaging. By integrating these 
complementary information together, we can see the addition of extra 
correlative image pairs can further eliminate ambiguous solutions in kernel 
estimation and image restoration. Furthermore, a prominent problem in multiple 
image deblurring is how well image pairs can be aligned. We then proposed a 
fully automatic alignment approach for multiple image pairs using sparseness 
prior of the blur kernel. Thus our methods are very practical and effective for 
achieving satisfactory photos in dim light conditions using off-of-shelf 
hand-held camera. The second idea is from the prior of sharp image structure. 
In our image deblurring with blurred/noisy image pair, the noisy image provides 
large-scale and sharp structures for accurate kernel estimation and 
high-quality image restoration as the guide image. I will show this insight can 
be furthermore applied to single image deblurring. As we observed, the 
reconstructed image usually contains unpleasant artifacts, i.e. ringing, due to 
the ill-posedness of the deconvolution even if the blur kernel is known. To 
suppress ringing artifacts and preserve restored structures, we require the 
guide image to tell where edges and texture regions are, and where flat regions 
are. Thus, we develop an inter-scale and intra-scale deconvolution framework to 
progressively recover such a guide image, which is used to adaptively suppress 
artifacts in texture regions and flat regions. Our progressive deconvolution 
approach can produce very promising results not only in synthetic experiments, 
but also in various types of real cases. Our results show our approach 
outperforms other state-of-the-art techniques and have wide applications in 
scientific and daily areas.


Date:     		Monday, 22 June 2009

Time:                   3:30pm-5:30pm

Venue:                  Room 4483
 			lifts 25-26

Committee Members:      Prof. Long Quan (Supervisor)
 			Dr. Chi-Keung Tang (Chairperson)
 			Dr. Huamin Qu
 			Dr. Chiew-Lan Tai


**** ALL are Welcome ****