Title

Deep Richardson-Lucy Deconvolution

DGIST Authors

Jo, Janghun ; Cho, Sunghyun ; Chwa, Hoon Sung

Advisor

좌훈승

Co-Advisor(s)

Sunghyun Cho

Issued Date

2020

Awarded Date

2020-02

Citation

Janghun Jo. (2020). Deep Richardson-Lucy Deconvolution. doi: 10.22677/Theses.200000283399

Type

Thesis

Description

image processing, deblurring, deep learning

Table Of Contents

I. Introduction 1
II. Background 4
2.1. Non-blind Image Deblurring 4
2.2. Richardson-Lucy Deconvolution 5
III. Related Work 6
3.1. Non-uniform blur. 6
IV. Methods 8
4.1. Parameterizing Richardson-Lucy Deconvolution 8
4.2. Deep Richardson-Lucy Deconvolution 9
4.3. Feature Scaling 11
4.4. Non-uniform Deblurring 11
4.5. Handling Boundary Artifact 13
4.6. Training 14
4.7. Dataset 14
4.8. Implementation 15
V. Experiments 17
5.1. Uniform Deblurring 17
5.2. Non-uniform Deblurring 20
5.3. Evaluation on Poisson noise 22
VI. Conclusion 24
References 25

URI

http://dgist.dcollection.net/common/orgView/200000283399
http://hdl.handle.net/20.500.11750/12021

DOI

10.22677/Theses.200000283399

Degree

Master

Department

Information and Communication Engineering

Publisher

DGIST

Show Full Item Record