메뉴 건너뛰기

SuperResolution Image Processing Lab.

본 연구실 졸업생인 김한솔 박사가 연구한 “Self-learning based joint multi image super-resolution and sub-pixel registration” 제목의 논문이 Digital Signal Processing (Impact factor : 2.9, Rank : Q2)에 게재가 확정되었다. 본 연구는 김한솔 박사(1저자)와 동서대학교 정보통신공학과 이석호 교수(2저자), 강문기 교수(교신저자)가 진행하였다.


Abstract

Multi Image Super-resolution (MISR) refers to the task of enhancing the spatial resolution of a stack of lowresolution (LR) images representing the same scene. Although many deep learning-based single image superresolution (SISR) technologies have recently been developed, deep learning has not been widely exploited for

MISR, even though it can achieve higher reconstruction accuracy because more information can be extracted

from the stack of LR images. One of the primary obstacles encountered by deep networks when addressing the

MISR problem is the variability in the number of LR images that act as input to the network. This impedes the

feasibility of adopting an end-to-end learning approach, because the varying number of input images makes

it difficult to construct a training dataset for the network. Another challenge arises from the requirement to

align the LR input images to generate high-resolution (HR) image of high quality, which requires complex and

sophisticated methods.

In this paper, we propose a self-learning based method that can simultaneously perform super-resolution and

sub-pixel registration of multiple LR images. The proposed method trains a neural network with only the LR

images as input and without any true target HR images; i.e., the proposed method requires no extra training

dataset. Therefore, it is easy to use the proposed method to deal with different numbers of input images. To

our knowledge this is the first time that a neural network is trained using only LR images to perform a joint

MISR and sub-pixel registration. Experimental results confirmed that the HR images generated by the proposed

method achieved better results in both quantitative and qualitative evaluations than those generated by other

deep learning-based methods.

번호 제목 글쓴이 날짜 조회 수
» 김한솔 박사 Self-learning based joint multi image super-resolution and sub-pixel registration 논문 Digital Signal Processing 게재 웹마스터 2024.11.25 59
84 이진욱 학생 Blind Image Deblurring via Bayesian Estimation Using Expected Loss 논문 IEEE Access 게재 웹마스터 2024.10.02 56
83 정경훈 박사 Multispectral Demosaicing Based on Iterative-Linear-Regression Model for Estimating Pseudo-Panchromatic Image 논문 Sensors 게재 웹마스터 2024.10.02 61
82 김한솔 박사 Multi-frame demosaicing for the Quad Bayer CFA in the color difference domain 논문 Optics Express 게재 웹마스터 2024.10.02 40
81 이해근 박사 Overlapping Group Prior Image Deconvolution논문 SIGNAL PROCESSING 게재 관리자 2023.05.22 304
80 창립기념일 "최우수 피인용 표창" 수상 file 웹마스터 2023.05.15 223
79 박종은 학생 Super-Resolution Kernel Estimation 논문 Sensors에 게재 관리자 2023.04.07 239
78 이해근 박사 Infrared Image Deconvolution논문 Sensors에 게재 관리자 2023.03.13 174
77 홍순영 학생 Single image dehazing논문 Neurocomputing에 게재 관리자 2021.12.17 430
76 이성우 학생 신호처리합동학술대회 우수논문 선정 관리자 2021.10.01 285
75 이해근 학생 Automatic Prior Selection 논문 Signal Processing에 게재 관리자 2021.08.30 4559
74 한재덕 박사 Thermal Image Restoration 논문 Sensors에 게재 관리자 2021.08.30 411
73 이상윤 학생 Poisson Gaussian noise reduction 논문 IEEE Access에 게재 관리자 2021.08.10 1119
72 홍순영 학생 Nighttime image dehazing논문 IEEE Access에 게재 관리자 2021.08.10 212
71 김민섭 학생 Road scene dehazing/adaptive atmospheric PSF논문 IEEE Access에 게재 관리자 2021.08.10 15479
70 김종현 학생 Depth Super-resolution 연구, IEEE Access 논문 게재 file 관리자 2020.09.07 1068
69 홍순영 학생 Signal Processing 에 논문게재 확정 조교 2020.08.11 2335
68 김한솔 학생 Sensors 에 논문게재 확정 관리자 2020.08.11 35697
67 한재덕 학생 IEEE Transactions on Circuits and Systems for video Technology 에 논문게재 확정 file 관리자 2020.03.03 1991
66 송기선 연구원 IEEE Transactions on Circuits and Systems for video Technology 에 논문게재 확정 file 관리자 2019.11.02 3498