Image Non-blind Deconvolution
[Click to download
a package with the executable and examples]
This is an executable for non-blind deconvolution.
The command is:
robust_deconv.exe inputImageName kernelFileName
outputImageName boundaryType noiseStr localPriorWeight
Parameter Definition
|
inputImageName
|
The
filename of the input blur image; .bmp, .png and .jpg files are
supported |
|
kernelFileName
|
The filename of the kernel image |
|
outputImageName
|
The filename of the target image which is used to store
the result |
|
boundaryType
|
The boundary type.
If boundaryType = 1, our program automatically
handles the boundary problem as described in our paper,
which takes some time. If boundaryType
= 0, the boundary condition is not modified and the boundary ringing artifacts
may be caused.
|
| noiseStr |
A parameter indicating the strength of noise. A very
large value may smooth the result, while a small
parameter helps reveal more fine details but possibly produces more visual artifacts.
The range is [0.001, 0.2]. |
| localPriorWeight |
A
local prior weight
parameter to adjust the smoothness enforced in flat
regions.
Its range is [1, 100]. |
System Requirement
Windows XP, 2003 Server, or Vista.
Version Information
Distribution Version: 0.1 (Nov. 6th, 2008) by Qi Shan and Jiaya
Jia ({qshan,
leojia}@cse.cuhk.edu.hk), CSE department, the Chinese
University of Hong Kong. This program is tested on Windows XP, 2003
Server and Vista, but is still not guaranteed to be
bug-free and work properly with all versions of Windows. It is for
educational use ONLY. If you use this executable for your academic
publication, please acknowledge this work:
@article{hqdeblurring_siggraph2008,
author = {Qi Shan and Jiaya Jia and Aseem Agarwala},
title = {High-quality Motion Deblurring from a Single Image},
journal = {ACM Transactions on Graphics (SIGGRAPH)},
year = {2008},
}
Examples:
Input Image
|
Output Image
|
Kernel
|
 |
 |
noiseStr = 0.02
localPriorWeight = 1
|
 |
 |
noiseStr = 0.008
localPriorWeight = 20
|
 |
 |
noiseStr = 0.04
localPriorWeight = 50 |
