• Journal of the Korea Computer Graphics Society
• /
• v.28 no.3
• /
• pp.55-65
• /
• 2022

Recently, deep learning-based automated systems for identifying and detecting landmarks have been proposed. In order to train such a deep learning-based model without overfitting, a large amount of image and labeling data is required. Conventionally, an experienced reader manually identifies and labels landmarks in a patient's image. However, such measurement is not only expensive, but also has poor reproducibility, so the need for an automated labeling method has been raised. In addition, in the X-ray image, since various human tissues on the path through which the photons pass are displayed, it is difficult to identify the landmark compared to a general natural image or a 3D image modality image. In this study, we propose a geometric data augmentation technique that enables the generation of a large amount of labeling data in X-ray images. In addition, the optimal attention mechanism for landmark detection was presented through the implementation and application of various attention techniques to improve the detection performance of 16 major landmarks in the skull. Finally, among the major cranial landmarks, markers that ensure stable detection are derived, and these markers are expected to have high clinical application potential.

• Journal of Space Technology and Applications
• /
• v.2 no.2
• /
• pp.104-120
• /
• 2022

The Small Scale magNetospheric and Ionospheric Plasma Experiment (SNIPE)'s scientific goal is to observe spatial and temporal variations of the micro-scale plasma structures on the topside ionosphere. The four 6U CubeSats (~10 kg) will be launched into a polar orbit at ~500 km. The distances of each satellite will be controlled from 10 km to more than ~1,000 km by the formation flying algorithm. The SNIPE mission is equipped with identical scientific instruments, Solid-State Telescopes(SST), Magnetometers(Mag), and Langmuir Probes(LP). All the payloads have a high temporal resolution (sampling rates of about 10 Hz). Iridium communication modules provide an opportunity to upload emergency commands to change operational modes when geomagnetic storms occur. SNIPE's observations of the dimensions, occurrence rates, amplitudes, and spatiotemporal evolution of polar cap patches, field-aligned currents (FAC), radiation belt microbursts, and equatorial and mid-latitude plasma blobs and bubbles will determine their significance to the solar wind-magnetosphere-ionosphere interaction and quantify their impact on space weather. The formation flying CubeSat constellation, the SNIPE mission, will be launched by Soyuz-2 at Baikonur Cosmodrome in 2023.

• Journal of Internet Computing and Services
• /
• v.22 no.6
• /
• pp.129-140
• /
• 2021

The next generation mBcN should be built to cooperate with the wireless network to support hyper-speed and hyper-connectivity. In this paper, we propose a network architecture for the cooperation mBcN and 5G commercial network and architecture of the cooperative security gateway required for the cooperation. The proposed cooperative security gateway is between gNB and UPF to support LBO, SFC, and security. Our analysis shows that the proposed architecture has several advantages. First of all, user equipment connected with the mBcN can be easily connected through the 5G commercial radio network to the mBcN. Second, the military application traffic can be transmitted to mBcN without going through the 5G core network, reducing the end-to-end transmission delay without causing the traffic load on the 5G core network. In addition, the security level of the military application can effectively be maintained because the user equipment can be connected to the cooperative security gateway, and the traffic generated by the user equipment is transmitted to the mBcN without going through the 5G core network. Finally, we demonstrate that LBO, SFC, and security modules are essential functions of the proposed gateway in the 5G test-bed environment.

• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.1-16
• /
• 2022

Despite recent breakthroughs in deep learning and computer vision fields, the pixel-wise identification of tiny objects in high-resolution images with complex disturbances remains challenging. This study proposes a modified U-net for tiny crack segmentation in real-world steel-box-girder bridges. The modified U-net adopts the common U-net framework and a novel Self-Attention-Self-Adaption (SASA) neuron as the fundamental computing element. The Self-Attention module applies softmax and gate operations to obtain the attention vector. It enables the neuron to focus on the most significant receptive fields when processing large-scale feature maps. The Self-Adaption module consists of a multiplayer perceptron subnet and achieves deeper feature extraction inside a single neuron. For data augmentation, a grid-based crack random elastic deformation (CRED) algorithm is designed to enrich the diversities and irregular shapes of distributed cracks. Grid-based uniform control nodes are first set on both input images and binary labels, random offsets are then employed on these control nodes, and bilinear interpolation is performed for the rest pixels. The proposed SASA neuron and CRED algorithm are simultaneously deployed to train the modified U-net. 200 raw images with a high resolution of 4928 × 3264 are collected, 160 for training and the rest 40 for the test. 512 × 512 patches are generated from the original images by a sliding window with an overlap of 256 as inputs. Results show that the average IoU between the recognized and ground-truth cracks reaches 0.409, which is 29.8% higher than the regular U-net. A five-fold cross-validation study is performed to verify that the proposed method is robust to different training and test images. Ablation experiments further demonstrate the effectiveness of the proposed SASA neuron and CRED algorithm. Promotions of the average IoU individually utilizing the SASA and CRED module add up to the final promotion of the full model, indicating that the SASA and CRED modules contribute to the different stages of model and data in the training process.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.42 no.3
• /
• pp.351-357
• /
• 2022

A conventional seawater reverse osmosis (SWRO) and a forward osmosis (FO) and reverse osmosis (RO) hybrid process to produce 1,000 m³/d of fresh water, were designed and compared in terms of carbon emission. When FO was adapted for the osmotic dilution, the required pressure for RO decreases, and thus energy consumption decreases. The decrease in carbon emission by decreased energy consumption (up to -0.73 kgCO₂/m³ using coal as the energy source) was compared with the increase in carbon emission by the FO system (+0.16 kgCO₂/m³), which is a function of various factors such as the number of FO modules and energy consumption. The comparison revealed that the FO-RO process causes less carbon emission compared with the SWRO process when the energy sources are coal and oil. However, if energy sources with low carbon emission such as solar, wind, and nuclear energy are selected, the carbon emission of the FO-RO process becomes higher than that of the SWRO process. This implies that the type of energy source is a key factor to determine the necessity of the FO-RO process from the aspect of carbon emission.

• Journal of the Korean Society of Radiology
• /
• v.16 no.1
• /
• pp.25-34
• /
• 2022

Brain computed tomography (CT) is useful for brain lesion diagnosis, such as brain hemorrhage, due to non-invasive methodology, 3-dimensional image provision, low radiation dose. However, there has been numerous misdiagnosis owing to a lack of radiologist and heavy workload. Recently, object detection technologies based on artificial intelligence have been developed in order to overcome the limitations of traditional diagnosis. In this study, the applicability of a deep learning-based YOLOv5s model was evaluated for brain hemorrhage detection using brain CT images. Also, the effect of hyperparameters in the trained YOLOv5s model was analyzed. The YOLOv5s model consisted of backbone, neck and output modules. The trained model was able to detect a region of brain hemorrhage and provide the information of the region. The YOLOv5s model was trained with various activation functions, optimizer functions, loss functions and epochs, and the performance of the trained model was evaluated in terms of brain hemorrhage detection accuracy and training time. The results showed that the trained YOLOv5s model is able to provide a bounding box for a region of brain hemorrhage and the accuracy of the corresponding box. The performance of the YOLOv5s model was improved by using the mish activation function, the stochastic gradient descent (SGD) optimizer function and the completed intersection over union (CIoU) loss function. Also, the accuracy and training time of the YOLOv5s model increased with the number of epochs. Therefore, the YOLOv5s model is suitable for brain hemorrhage detection using brain CT images, and the performance of the model can be maximized by using appropriate hyperparameters.

• Journal of Korean Library and Information Science Society
• /
• v.53 no.3
• /
• pp.431-454
• /
• 2022

This study examined data science degree programs and courses offered by universities, and those offered by the Library and Information Science (LIS) degree programs, to understand its implications for the LIS programs in Korea. This research assessed the status of data science degrees from 439 schools using the list released by the Korea Educational Development Institute in 2022. To be specific, this study analyzed universities, colleges, majors, sub-majors, interdisciplinary majors, convergence majors, micro-degrees, nanodegrees, tracks, modules, and industry-university cooperative programs within the data science field. This research examined 1,148 courses offered by data science degree programs and 1,325 courses offered by LIS degree programs. Data science degrees in Korea offer courses such as introductory, technical, practical, applied, and in-depth subjects related to data science. Although the LIS programs in Korea do not always offer data science, the courses included topics such as the introduction to data science, database, data visualization, data curation, metadata, big data, and information technology, when courses were offered. The researchers hope the findings of this study will be useful as a starting point for the development and revisions of LIS curriculum on data science in Korea.

• Geophysics and Geophysical Exploration
• /
• v.25 no.3
• /
• pp.129-139
• /
• 2022

Analog-based streamers for ultra-high-resolution seismic surveys are capable of additional noise ingress in water, but the specifications cannot be expanded through interconnections. Foreign-produced digital streamers have been introduced and used primarily at domestic research institutes; however, the cost is high and smooth maintenance is challenging. This study investigates the localization of ultra-high-resolution digital streamers capable of high-resolution imaging of a geological structure. A digital streamer capable of 24-bit, 10 kHz digital sampling of up to 64 channel data was developed through research and development. Various quantitative specifications of the system were designed and developed close to the benchmark model, Geometrics' GeoEel streamer, and the number of modules that make up the system was drastically reduced, reducing development costs and making it easier to use. The field applicability of the developed streamer system was evaluated in an in situ experiment conducted in the waters around the Port of Yeong-il Bay in Pohang in April 2022.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.23 no.2
• /
• pp.157-162
• /
• 2023

Among the failures of photovoltaic power generation facilities, failures caused by surges account for 20% of the total failure rate, and energy emissions of tens to hundreds [A] during power generation and electrical damage to inverters and connection boards lead to electrical safety accidents. In particular, in the case of lightning, an abnormal voltage is induced in an electric circuit to destroy insulation, and the current flowing at this time causes a fire and acts as a factor that accelerates the deterioration of parts. Due to this action, the problem of electrical safety of solar power generation devices spreading from outside the city center to the inside of the city center such as houses, apartments, and government offices is emerging. Since lightning strikes cause both field-based and conducted electrical interference, this effect increases with increasing cable length or conductor loops. In addition, surge damages not only solar modules, inverters and monitoring devices, but also building facilities, which can eventually cause operational shutdown due to fire of the photovoltaic power generation system and consequent financial loss. Therefore, in this paper, a lightning protection system for solar power generation devices is studied for the purpose of reducing property damage and human casualties due to the increase in fire and electrical safety accidents caused by lightning strikes in photovoltaic power generation systems.

• Journal of Industrial Convergence
• /
• v.20 no.11
• /
• pp.19-25
• /
• 2022

In the development of existing games, it is judged that virtual world production was done by arranging game elements one by one. What is noteworthy here is the question of whether quality virtual worlds were efficiently produced in preparation for investment. In this study, we propose a methodology that can build an efficient virtual world based on the concept of modularization in an unreal engine. First, precedents were analyzed and five reference elements for modularization were extracted. In addition, the concept of an instance production pipeline was proposed by dividing it into four stages, and the minimum-unit instance modules for urban virtual world production were compressed into four. Finally, an urban virtual world constructed based on the minimum unit module and reference elements was implemented and presented. In conclusion, research on the production method centered on this efficiency is thought to be able to focus the time that designers or artists had to spend on production only on ideas and creativity. The limitations of the research are that the basic minimum module is limited to the city, and the derived reference elements and production pipelines have not been verified when implementing them with an unreal engine. Therefore, it is expected that various virtual world creation plans will be derived through more advanced modular research.