DaeSoon Park

Mr. DaeSoon Park

orcid

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
8 October 2020 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Closed access
JSJoonyoung Song
Joonyoung Song
JJJae-Heon Jeong
Jae-Heon Jeong
DPDaeSoon Park
DaeSoon Park
HKHyun-Ho Kim
Hyun-Ho Kim
DSDoo-Chun Seo
Doo-Chun Seo
JYJong Chul Ye
Jong Chul Ye
Multispectral satellite imaging sensors acquire various spectral band images and have a unique spectroscopic property in each band. Unfortunately, image artifacts from imaging sensor noise often affect the quality of scenes and have a negative impact on applications for satellite imagery. Recently, deep learning approaches have been extensively explored to remove noise in satellite imagery. Most deep learning denoising methods, however, follow a supervised learning scheme, which requires matched noisy image and clean image pairs that are difficult to collect in real situations. In this article, we propose a novel unsupervised multispectral denoising method for satellite imagery using a wavelet directional cycle-consistent adversarial network (WavCycleGAN). The proposed method is based on an unsupervised learning scheme using adversarial loss and cycle-consistency loss to overcome the lack of paired data. Moreover, in contrast to the standard image-domain cycleGAN, we introduce a wavelet directional learning scheme for effective denoising without sacrificing high-frequency components such as edges and detailed information. Experimental results for the removal of vertical stripes and wave noise in satellite imaging sensors demonstrate that the proposed method effectively removes noise and preserves important high-frequency features of satellite images.

12 December 2024 in Remote Sensing
Verified
Journal Article
Open access
PDF file
DLDongHan Lee
DongHan Lee
DPDaeSoon Park
DaeSoon Park
DYDaehoon Yoo
Daehoon Yoo
After the launch of a high-resolution remote sensing satellite, representative spatial quality estimators (RER, FWHM, MTF50, MTFA) are measured from images taken of ground Edge targets. In this work, the best spatial quality estimator is proposed by quantitatively comparing and analyzing the precision between the Relative Edge Response (RER), the Full Width at Half Maximum (FWHM), the MTF value at the Nyquist frequency (MTF50), and the MTF Area between 0 and the Nyquist frequency (MTFA). While the basic method for the measurement of spatial quality estimators on Edge targets is already well established, this work summarizes and explains the uncertain factors and problems in the measurement procedure that affect the accuracy and precision of spatial quality estimators. It also considers how to improve the precision of spatial quality estimators during the measurement procedure. The contents and results of this work were discussed by various satellite development organizations in the Geo-Spatial Working Group within CEOS WGCV IVOS from 2012 to 2019, and the Edge target Spatial quality Measurement Python code (ESMP) was developed in 2019 to reflect the findings of this workshop. Using 483 Edge targets from worldwide images taken by KOMPSAT-3A, which has been in operation since 2017, the results obtained via ESMP show that the precision levels of RER, FWHM, and MTFA are approximately three to four times higher than that of MTF50 when comparing the Coefficient of Variance (CV) statistics. This is the first statistical comparison of spatial quality estimators using 7 years of ground Edge target imagery from a single satellite of KOMPSAT-3A.

22 October 2024 in Preprints.org
Verified
Preprint
Open access
PDF file
DLDongHan Lee
DongHan Lee
DPDaeSoon Park
DaeSoon Park
DYDaehoon Yoo
Daehoon Yoo
After the launch of a high-resolution remote sensing satellite, representative spatial quality estimators (RER, FWHM, MTF50, MTFA) are measured from images taken of ground Edge targets. In this work, the best spatial quality estimator is proposed by quantitatively comparing and analyzing the precision between Relative Edge Response (RER), Full Width at Half Maximum (FWHM), MTF value at Nyquist frequency (MTF50), and MTF Area between 0 and Nyquist frequency (MTFA). While the basic method for the measurement of spatial quality estimators on Edge targets is already well established, this work summarizes and explains the uncertain factors and problems in the measurement procedure that affect the accuracy and precision of spatial quality estimators. It also considers how to improve the precision of spatial quality estimators during the measurement procedure. The contents and results of this work were discussed by various satellite development organizations in the Geo-Spatial Working Group within CEOS WGCV IVOS from 2012 to 2019, and the Edge target Spatial quality Measurement Python code (ESMP) was developed in 2019 to reflect the findings of this workshop. Using 483 Edge targets from worldwide images taken by KOMPSAT-3A, which has been in operation since 2017, the results obtained via ESMP show that the precision of RER, FWHM, and MTFA are approximately three to four times higher than that of MTF50 when comparing the Coefficient of Variance (CV) statistics.

23 February 2020 in arXiv
Verified
Preprint
Open access
PDF file
JSJoonyoung Song
Joonyoung Song
JJJae-Heon Jeong
Jae-Heon Jeong
DPDaeSoon Park
DaeSoon Park
HKHyun-Ho Kim
Hyun-Ho Kim
DSDoo-Chun Seo
Doo-Chun Seo
JYJong Chul Ye
Jong Chul Ye
Multi-spectral satellite imaging sensors acquire various spectral band images such as red (R), green (G), blue (B), near-infrared (N), etc. Thanks to the unique spectroscopic property of each spectral band with respective to the objects on the ground, multi-spectral satellite imagery can be used for various geological survey applications. Unfortunately, image artifacts from imaging sensor noises often affect the quality of scenes and have negative impacts on the applications of satellite imagery. Recently, deep learning approaches have been extensively explored for the removal of noises in satellite imagery. Most deep learning denoising methods, however, follow a supervised learning scheme, which requires matched noisy image and clean image pairs that are difficult to collect in real situations. In this paper, we propose a novel unsupervised multispectral denoising method for satellite imagery using wavelet subband cycle-consistent adversarial network (WavCycleGAN). The proposed method is based on unsupervised learning scheme using adversarial loss and cycle-consistency loss to overcome the lack of paired data. Moreover, in contrast to the standard image domain cycleGAN, we introduce a wavelet subband domain learning scheme for effective denoising without sacrificing high frequency components such as edges and detail information. Experimental results for the removal of vertical stripe and wave noises in satellite imaging sensors demonstrate that the proposed method effectively removes noises and preserves important high frequency features of satellite images.