Home // SIGNAL 2016, The First International Conference on Advances in Signal, Image and Video Processing // View article

Super-resolving a Single Blurry Image Through Blind Deblurring Using ADMM

Authors:
Xiangrong Zeng
Yan Liu
Jun Fan
Qizi Huangpeng
Jing Feng
Jinlun Zhou
Maojun Zhang

Keywords: Image super-resolution; Blind image deblurring; ADMM.

Abstract:
Both single blind image super-resolution (SBISR) and blind image deblurring (BID) are ill-posed inverse problems typically addressed by imposing some form of regularization (prior knowledge) on the unknown blurs and original images (the high resolution image and the sharp image for SBISR and BID, respectively). However, SBISR is more ill-posed than BID due to the introduce of the downsampling operator in the former, thus, the latter is usually easier to be solved than the former. We propose to address the SBISR problem by a BID method via reformulating it into a BID problem by an interpolation operator, and then solving the BID problem using the alternating direction method of multipliers (ADMM). Our approach bridges the gap between SBISR and BID, taking advantages of existing BID methods to handle SBISR. Experiments on synthetic and real blurry images (also on a real sharp image) show that the proposed method is effective, and competitive in terms of speed and restoration quality.

Pages: 1 to 5

Copyright: Copyright (c) IARIA, 2016

Publication date: June 26, 2016

Published in: conference

ISSN: 2519-8432

ISBN: 978-1-61208-487-9

Location: Lisbon, Portugal

Dates: from June 26, 2016 to June 30, 2016