Search

Enhancement of vessel detection performance in synthetic aperture radar (SAR) images generated academic advancements related to amelioration of the algorithmic accuracy and training data procurement. For practical implementation of vessel detection algorithm to maritime surveillance, however, presentation of authentic position of vessels was essential. Accordingly, this article aimed to propose an algorithm, which demonstrated realistic and azimuth shift-corrected position of vessel, especially out of conventional vessel monitoring apparatus: automated identification system (AIS) and vessel-pass (VPASS) information. Two different analyses regarding the vessel detection output utilization were, therefore, presented. Primary analysis demonstrated a vessel identification algorithm, comparing the vessel detection output with elaborately preprocessed AIS and VPASS information, which indicated the discrete position and velocity of vessel. The other presented a position restoration algorithm via i) velocity estimator complementing the conventional fractional Fourier transform velocity estimation analysis, while investigating the effect of range acceleration in deriving the azimuth velocity and ii) measuring the vessel orientation angle from Radon transform. Both algorithms were implemented to the vessel detection output in Cosmo-SkyMed SAR images, depicting an enhanced accuracy compared to the conventional algorithm both in velocity estimation and azimuth shift estimation; velocity offset reduced from 1.64 m/s to 1.29 m/s and average azimuth shift offset reduced from 70.75 m to 62.39 m. The presented algorithms would be decisive in terms of practicality if robustly attached to convolutional neural network-based vessel detection algorithms demonstrating ideal detection performances.

Some point out that the influence of YouTube's video recommendation algorithm is causing users to be exposed to only video clips in limited subjects or fields, especially to biased content with opinions that are tilted to one side. However, there is a lack of empirical research on filter bubbles as algorithms in YouTube have not been disclosed. This study indirectly demonstrated the phenomenon of filter bubble on YouTube by extracting comment-based content network between uploaders who posted videos and writers who wrote comments on the video by each subject of the contents. Also, this study analyzed communication patterns between users through social network analysis (SNA). According to the analysis, users' narrow information acquisition and communication phenomenon caused by the filter bubble in YouTube was found.

This manuscript presents a self-interferometric phase analysis technique for sea surface observation using a single scatterometer system. The self-interferometric phase is proposed to complement the imprecise analysis results due to the very meager signal strength measured at a high incident angle of more than 30°, which is a vulnerability of the existing analysis method using the Doppler frequency based on the backscattered signal strength. Moreover, compared to conventional interferometry, it is characterized by the phase-based analysis using consecutive signals from a single scatterometer system without any auxiliary system or channel. To apply the interferometric signal process on the moving sea surface observation, it is necessary to secure a reference target; however, this is hard to solve in practice. Hence, we adopted the back-projection algorithm to project the radar signals onto a fixed reference position above the sea surface, where the theoretical model for extracting the self-interferometric phase was derived from the radar-received signal model applying the back-projection algorithm. The observation performance of the proposed method was verified using the raw data collected at the Ieodo Ocean Research Station in Republic of Korea. In the observation result for wind velocity at the high incident angles of 40° and 50°, the self-interferometric phase analysis technique shows a better performance of a correlation coefficient of more than about 0.779 and an RMSE (root-mean-square error) of about 1.69 m/s compared to the existing method of a correlation coefficient of less than 0.62 and RMSE of more than 2.46 m/s.

Estimating horizontal winds in and around typhoons is important for improved monitoring and prediction of typhoons and mitigating their damages. Here, we present a new algorithm for estimating typhoon winds using multiple satellite observations and its application to Typhoon Soulik (2018). Four kinds of satellite remote sensing data, along with their relationship to typhoon intensity, derived statistically from hundreds of historical typhoon cases, were merged into the final product of typhoon wind (MT wind): 1) geostationary-satellite-based infrared images (IR wind), 2) passive microwave sounder (MW wind), 3) feature-tracked atmospheric motion vectors, and 4) scatterometer-based sea surface winds (SSWs). The algorithm was applied to two cases (A and B) of Typhoon Soulik and validated against SSWs independently retrieved from active microwave synthetic aperture radar (SAR) and microwave radiometer (AMSR2) images, and vertical profiles of wind speed derived from reanalyzed data and dropsonde observations. For Case A (open ocean), the algorithm estimated the realistic maximum wind, radius of maximum wind, and radius of 15 m/s, which could not be estimated using the reanalysis data, demonstrating reasonable and practical estimates. However, for Case B (when the typhoon rapidly weakened just before making landfall in the Korean Peninsula), the algorithm significantly overestimated the parameters, primarily due to the overestimation of typhoon intensity. Our study highlights that realistic typhoon winds can be monitored continuously in real-time using multiple satellite observations, particularly when typhoon intensity is reasonably well predicted, providing timely analysis results and products of operational importance.

Human civilization is rapidly developing thanks to the spread of geospatial technology and artificial intelligence [...]

Satellite-based flood monitoring for providing visual information on the targeted areas is crucial in responding to and recovering from river floods. However, such monitoring for practical purposes has been constrained mainly by obtaining and analyzing satellite data, and linking and optimizing the required processes. For these purposes, we present a deep learning-based flood area extraction model for a fully automated flood monitoring system, which is designed to continuously operate on a cloud-based computing platform for regularly extracting flooded area from Sentinel-1 data, and providing visual information on flood situations with better image segmentation accuracy. To develop the new flood area extraction model using deep learning, initial model tests were performed more than 500 times to determine optimal hyperparameters, water ratio, and best band combination. The results of this research showed that at ‘waterbody ratio 30%’, which yielded higher segmentation accuracies and lower loss, precision, overall accuracy, IOU, recall, and F1 score of ‘VV, aspect, topographic wetness index, and buffer input bands’ were 0.976, 0.956, 0.894, 0.964, and 0.970, respectively, and averaged inference time was 744.3941 s, which demonstrate improved image segmentation accuracy and reduced processing time. The operation and robustness of the fully automated flood monitoring system were demonstrated by automatically segmenting 12 Sentinel-1 images for the two major flood events in Republic of Korea during 2020 and 2022 in accordance with the hyperparameters, waterbody ratio, and band combinations determined through the intensive tests. Visual inspection of the outputs showed that misclassification of constructed facilities and mountain shadows were extremely reduced. It is anticipated that the fully automated flood monitoring system and the deep leaning-based waterbody extraction model presented in this research could be a valuable reference and benchmark for other countries trying to build a cloud-based flood monitoring system for rapid flood monitoring using deep learning.

Sea surface temperature (SST) is crucial for understanding the physical characteristics and ecosystems of coastal seas. SST varies near the tidal flat, where exposure and flood recur according to the tidal cycle. However, the variability of SST near the tidal flat is poorly understood owing to difficulties in making in-situ observations. The high resolution of Landsat 8 enabled us to determine the variability of SST near the macro tidal flat. The spatial distribution of the SST extracted from Landsat 8 changed drastically. The seasonal SST range was higher near the tidal flat than in the open sea. The maximum seasonal range of coastal SST exceeded 23 °C, whereas the range in the open ocean was approximately 18 °C. The minimum and maximum horizontal SST gradients near the tidal flat were approximately − 0.76 °C/10 km in December and 1.31 °C/10 km in June, respectively. The heating of sea water by tidal flats in spring and summer, and cooling in the fall and winter might result in a large horizontal SST gradient. The estimated heat flux from the tidal flat to the seawater based on the SST distribution shows seasonal change ranging from − 4.85 to 6.72 W/m2.

Accurately tracking marine traffic considering security and commercial activities is still challenging despite its increasing global importance. Recently, space-borne synthetic aperture radar (SAR) is being considered to accurately monitor maritime traffic, and techniques to detect the position of ships and estimate their velocity have become essential. Here, we investigated the potential for automatic estimation of ship velocity using the azimuth offset between ships and wakes detected using convolutional neural network (CNN) coupled with SAR imagery. We found that azimuth offset is proportional to the Doppler shift effect of the back-scattered signal in SAR, thus, it relates to the radial velocity of a moving target. Consequently, we propose a method whereby a CNN is applied to automatically detect ship wakes from TanDEM-X data. In this method, ship velocity is calculated using the azimuthal distance (i.e., azimuth offset) between the stern of the detected ship and the vertex of the detected V-shape wake-determined as the intersection of two lines obtained through edge filtering and Radon transforms. The location and number of detected ships are then compared with an automatic identification system (AIS), and the calculated velocity of the ship is compared with the velocity obtained via along-track interferometry and AIS. Results show that our method automatically detects ships and wakes with accuracies of 91.0% and 93.2%, respectively, and estimates the velocity of ships with an accuracy of 0.13 m/s. This method is effective when wind velocities are not substantially higher than 5.5 m/s and ship velocities are not extremely low.

This paper suggests a novel and reliable method to detect water body from Sentinel-1 SAR satellite data using deep learning technique. There have been a lot of studies to extract water body from SAR images with deep learning. Although they achieved good performance, most of them used training data without guaranteeing good quality. In this study, land cover map generated by an official government agency were used for labelling ground truth data. After identifying the acquisition date of aerial photo used for generating the land cover map, vector polygons for river or reservoir were extracted and used as label data. This new method reduced producing time and cost to generate reliable training data. After training our deep learning model, it showed 0.874 of f1score. We also tested our deep learning model to the heavy rain season in Korea (August 2020) and successfully detected river flooding.

Monitoring of tidal creeks is important for safety in tidal flats and for understanding coastal changes due to sedimentation and erosion. The tidal creek is a small stream developed on tidal flats and requires very precise Digital Elevation Model (DEM) to detect it. Interferometric SAR (InSAR) technique was usually applied for generating DEM over inland regions. In tidal flat regions, low backscattering due to flatness and remnant water of the tidal flat prevents from generating high precision DEM. In this study, optimum InSAR baseline condition was investigated to generate such a precise DEM in the tidal flat. The long-baseline TanDEM-X and airborne InSAR system were found to be useful, and the tidal creek information (width and depth) was successfully extracted from the DEM that was generated from optimal baseline InSAR systems.

Monitoring and detecting ships via machine learning based algorithm were regarded efficient in martial and economic manners. As an algorithm regarding automated training data retrieval from SAR image was proposed, the identification of unclassified ships without AIS information could be raised as another challenging issue of ship surveillance. This study presented the effective identification algorithm of discerning unclassified ships from AIS information and the results of conventional ship detection based on machine learning. The accurately detected ships were selected from the conventional ship detection results, followed by the preprocessing of AIS information corresponding to the SAR images containing the detection results. Superposition of AIS information on accurate detection results was conducted and concluded the ships without AIS information as unclassified ships. From 3 Sentinel-1 SAR images, it obtained the average rate of identification as 85.67%. Additional research implementing the identification algorithm accompanied by rapidly acquired satellite or airborne SAR images could be effective in rendering a ship surveillance system with rapid response.

For open ocean environments, it is rare to find continuous, simultaneous air and sea observation records due to the challenges of instrument installation and maintenance. The Ieodo Ocean Research Station (Ieodo ORS), a remote ocean site located in the northern East China Sea with its harsh oceanic and atmospheric environment, provides a platform for the concurrent monitoring of air and sea environments. Since 2014, the Korea Hydrographic and Oceanographic Agency has run the “Ieodo ORS field trip program,” via which researchers are able to stay at the station for a week or more. This work reports technical lessons learned over 5 years from five Ieodo ORS research projects launched in 2016. Over the course of these projects, Ieodo ORS has monitored sea surface temperature, temperature and salinity in the water column, seawater pH, air pollutants, and solar radiation. The purpose of this paper is to facilitate the success of future research activities in similar environments by sharing our experiences and “best practices.”

Recently, interests in spreading the culture of sharing, especially at the social level, are increasing due to the income polarization, poverty, the rapid growth of civil society, and the increase of importance of public interest in private organizations. According to the Ministry of Government Administration and Home Affairs, Korea’s volunteer participation rate of 2015 has steadily increased to 114% compared to 2011, and according to the World Donation Index evaluation, the ranks in interest and participation in non-monetary sharing have increased to 75th in 2016 and 60th in 2018. Even though social interest is increasing, not many empirical studies on the contribution to talent donation have been made yet. In 2016, a survey was conducted on the change of donation attitude and recognition of donation. The major topics in the area for spreading the donation culture in general were the convenience of donation methods and the transparency of donor organizations. In order to solve the convenience of donation, the mobile and Web based platforms such as Wikipedia, Airbnb, Delivery nation, and Yogiyo, make interactions possible without constraints on place, conditions and the number of user to induce the public to participate in talent donation. In addition, block chain is the distributed computing-based technology of a data forgery and tamper proofing. One person cannot modify the data arbitrarily, but anyone can see the result of the data change. The reason that block chain technology is in the spotlight is because anyone can access the result of data changes. That is, the transparency of the transaction can be secured. In this thesis, we intend to apply these block chain’s characteristics to the talent donation platform.

Recruitment corruption makes people demand fair and objective hiring procedures. The development of IT technology has promoted the emergence of recruitment system based on AI which make interview conduct without human intervention. As companies can reduce costs and efforts through the AI recruitment system, more and more companies are using the system. As the result, the research for figuring out the user intention on the system is needed. We designed the research model to find factors affecting the intention to use of AI-based recruitment system utilizing constructs of TAM and TOE. Also the moderating effect by the number of employees in the corporation will be estimated.

Recently, interest in the spread of the culture of sharing has been increasing at a social level due to the deepening income polarization, poverty, and the rapid growth of civil society and the importance of public interest activities of private organizations. In 2016, the National Statistical Office conducted a survey on changes in donation attitudes and recognition of donations, items that took up a large part in the area for the spread of the overall donation culture were the convenience of donation methods and the enhancement of transparency of donation organizations. The existing donation system is a way for donors to donate directly to several donor organizations. However, this method has its disadvantage: it is difficult for donors to see how their donations are spent. To address the convenience of such donation methods, based on the platform that is the core of the fourth industrial era, it enables unrestricted interaction in the network environment such as mobile or PC, inducing open participation in talent donation to the general public. Blockchain is a data forgery/modulation prevention technology based on distributed computing technology. Because the block chain records continuously changing data on all participating nodes, it is impossible to arbitrarily manipulate the data by the operators of distributed nodes. Blockchain technology is in the spotlight because it can secure transparency in transactions. In this paper, to solve these problems, we try to introduce the block chain technology into the donation system.

Assessment of bridge deflection and serviceability is essential for structural condition and infrastructure maintenance. In this study, we aim to estimate the high-resolution displacements of bridges using X-band Cosmo-SkyMed and C-band Sentinel-1 A/B SAR data. We applied multi-temporal Interferometric analysis to Geobuksan bridge and Dolsan Bridge in the Yeosu region. In addition to the bridge displacements, the geolocation error of scatterers were corrected by estimating DEM error.

Ship detection in coastal regions accompanied by machine learning could be effective in economic and martial issues. However, conventional researches on ship detection mainly focused on modifying the training model itself instead of obtaining qualified training data. In order to ameliorate the training data in the aspect of quality and quantity, this research aims to directly obtain training data from AIS information. From discrete AIS information, interpolation on SAR acquisition time was conducted, followed by adjusting the Doppler frequency shift caused by each vessel's velocity. Training data was constructed from the adjusted position using internal location of AIS sensor on each type of vessel. Extracted training data by the proposed algorithm from Sentinel-1 images was tested with CNN model. As the detection performance of the extracted training data exceeded that from visual interpretation, this study concluded that qualified training data could be extracted from the proposed algorithm.

Coastal sea level changes measured by tide gauges are affected by the severity of coastal land motion. Subsidence due to natural and anthropogenic activities in the coastal regions were directly exacerbates the tide gauge records. In this study, we aim to monitor the coastal land motion at the tide gauges in Korean peninsula using multi-temporal InSAR techniques. We present the methodology that simply employs the selection of sequential and small temporal interferograms in the StaMPS-SBAS technique for increased coherent PS pixels over coastal regions. C-band Sentinel-1 SAR data from ascending and descending mode were acquired over Pohang and Yeosu regions to map the coastal land motion. The preliminary results showed significant increased number of initial PS pixel density than conventional StaMPS-SBAS technique, chosen by Amplitude-Dispersion Index (ADI) and revealed the subsidence about 2 cm/year in Pohang tide gauge and small uplift of 1.46 mm/year near Yeosu tide gauge.

Relative sea-level changes observed by tide gauges are commonly corrected for several components including crustal displacement, ocean dynamics, and vertical land motion. Vertical Land Motion (VLM) due to local land hydrology is a crucial component that observed as localized ground motion and varies with each tide gauges. Permanent GNSS stations are used to measure the VLM trend at tide gauges, however, only few tide gauges are equipped with collocated GNSS stations. Multi-temporal InSAR analysis provides ground displacements in both the spatial and temporal domains. Therefore, in our study, we applied the spaceborne Interferometric SAR technique to measure the local ground motion using Sentinel-1 SAR data. The Korean peninsula is surrounded by the East Sea/Sea of Japan, the Yellow Sea and the East China Sea have continuously monitoring tide gauges with a record length of more than 30 years. We acquire C-band Sentinel-1 SAR data (both ascending and descending mode) over the Korean Peninsula during 2014/11 and 2019/04. We estimate the high-resolution (~ 10 m) land motion at tide gauges (mm-level accuracy) over these 21 tide gauges and, compared with available collocated GNSS observations. 2D displacements (vertical and horizontal) are derived from ascending and descending mode InSAR displacements. The linear trend of VLM observed from our InSAR estimates is used to compensate for the relative velocity of sea-level changes observed from tide gauges.

Tidal flats which are located between land and ocean are very vulnerable to the change of environment. These places are now facing many environmental challenges related to climate change and human-induced impacts. Topographic changes related to sedimentation or erosion in tidal flats are the most evident sign of the environmental changes. The tidal flats usually have small variations of topographies and experience ebb and flood tides every day. The conventional SAR interferometric techniques (such as repeat-pass InSAR) cannot be applied for generating digital elevation models (DEMs) in these tidal flats because of low coherence caused by ebb and flood tides. In this study, we developed a long-baseline single-pass airborne interferometric SAR (InSAR) system that has a small ambiguity height. We collected InSAR data in several tidal flats, the west coast of Korean peninsula, using our airborne InSAR system and TanDEM-X. The TanDEM-X is a space borne quasi-single pass interferometic SAR system. We also investigated the expected accuracy of DEM as function of baseline and coherence. In general, longer baseline is better for generating very sensitive and fine DEM that is absolutely needed for tidal flats, because longer baseline has smaller ambiguity height. But longer baseline can also cause low coherence due to baseline decorrelation. As the baseline increase, common spectral region is also decreased resulting in poorer coherence and poor vertical accuracy. Based on these investigations, we proposed optimal baselines for airborne InSAR system and space-borne SAR system. Finally, we compared the constructed topographies from our long-baseline airborne InSAR and TanDEM-X with GPS-RTK measurements.