• Steel and Composite Structures
• /
• v.34 no.1
• /
• pp.107-122
• /
• 2020

Moment resisting stub columns (MRSCs) have increasingly adopted in special moment-resisting frame (SMF) systems in steel building structures, especially in Asian countries. The MRSCs typically provide a lower deformation capacity compared to shear-panel stub columns, a limited post-yield stiffness, and severe strength degradation as adopting slender webs. A new MRSC design with cored configuration, consisting of a core-segment and two side-segments using different steel grades, has been proposed in the study to improve the demerits mentioned above. Several full-scale components of the cored MRSC were experimentally investigated focusing on the hysteretic performance of plastic hinges at the ends. The effects of the depths of the core-segment and the adopted reduced column section details on the hysteretic behavior of the components were examined. The measured hysteretic responses verified that the cored MRSC enabled to provide early yielding, great ductility and energy dissipation, enhanced post-yield stiffness and limited strength degradation due to local buckling of flanges. A parametric study upon the dimensions of the cored MRSC was then conducted using numerical discrete model validated by the measured responses. Finally, a set of model equations were established based on the results of the parametric analysis to accurately estimate strength backbone curves of the cored MRSCs under increasing-amplitude cyclic loadings.

• Journal of radiological science and technology
• /
• v.45 no.2
• /
• pp.151-158
• /
• 2022

The Cardiac Gated Blood Pool (GBP) scintigram, a nuclear medicine imaging, calculates the left ventricular Ejection Fraction (EF) by segmenting the left ventricle from the heart. However, in order to accurately segment the substructure of the heart, specialized knowledge of cardiac anatomy is required, and depending on the expert's processing, there may be a problem in which the left ventricular EF is calculated differently. In this study, using the DeepLabV3 architecture, GBP images were trained on 93 training data with a ResNet-50 backbone. Afterwards, the trained model was applied to 23 separate test sets of GBP to evaluate the reproducibility of the region of interest and left ventricular EF. Pixel accuracy, dice coefficient, and IoU for the region of interest were 99.32±0.20, 94.65±1.45, 89.89±2.62(%) at the diastolic phase, and 99.26±0.34, 90.16±4.19, and 82.33±6.69(%) at the systolic phase, respectively. Left ventricular EF was calculated to be an average of 60.37±7.32% in the ROI set by humans and 58.68±7.22% in the ROI set by the deep learning segmentation model. (p<0.05) The automated segmentation method using deep learning presented in this study similarly predicts the average human-set ROI and left ventricular EF when a random GBP image is an input. If the automatic segmentation method is developed and applied to the functional examination method that needs to set ROI in the field of cardiac scintigram in nuclear medicine in the future, it is expected to greatly contribute to improving the efficiency and accuracy of processing and analysis by nuclear medicine specialists.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.12
• /
• pp.4420-4438
• /
• 2021

Image dehazing is an ill-posed problem which is far from being solved. Traditional image dehazing methods often yield mediocre effects and possess substandard processing speed, while modern deep learning methods perform best only in certain datasets. The haze removal effect when processed by said methods is unsatisfactory, meaning the generalization performance fails to meet the requirements. Concurrently, due to the limited processing speed, most dehazing algorithms cannot be employed in the industry. To alleviate said problems, a lightweight fast dehazing network based on a multiple scale-patch framework (MSP) is proposed in the present paper. Firstly, the multi-scale structure is employed as the backbone network and the multi-patch structure as the supplementary network. Dehazing through a single network causes problems, such as loss of object details and color in some image areas, the multi-patch structure was employed for MSP as an information supplement. In the algorithm image processing module, the image is segmented up and down for processed separately. Secondly, MSP generates a clear dehazing effect and significant robustness when targeting real-world homogeneous and nonhomogeneous hazy maps and different datasets. Compared with existing dehazing methods, MSP demonstrated a fast inference speed and the feasibility of real-time processing. The overall size and model parameters of the entire dehazing model are 20.75M and 6.8M, and the processing time for the single image is 0.026s. Experiments on NTIRE 2018 and NTIRE 2020 demonstrate that MSP can achieve superior performance among the state-of-the-art methods, such as PSNR, SSIM, LPIPS, and individual subjective evaluation.

• Journal of Broadcast Engineering
• /
• v.28 no.3
• /
• pp.293-301
• /
• 2023

Unstandardized medical data collection and management are still being conducted manually, and studies are being conducted to classify CT data using deep learning to solve this problem. However, most studies are developing models based only on the axial plane, which is a basic CT slice. Because CT images depict only human structures unlike general images, reconstructing CT scans alone can provide richer physical features. This study seeks to find ways to achieve higher performance through various methods of converting CT scan to 2D as well as axial planes. The training used 1042 CT scans from five body parts and collected 179 test sets and 448 with external datasets for model evaluation. To develop a deep learning model, we used InceptionResNetV2 pre-trained with ImageNet as a backbone and re-trained the entire layer of the model. As a result of the experiment, the reconstruction data model achieved 99.33% in body part classification, 1.12% higher than the axial model, and the axial model was higher only in brain and neck in contrast classification. In conclusion, it was possible to achieve more accurate performance when learning with data that shows better anatomical features than when trained with axial slice alone.

• Journal of KIISE:Information Networking
• /
• v.32 no.2
• /
• pp.220-235
• /
• 2005

Internet Service Quality has been constantly the center of attention to Internet Service Providers. The KII project plan (Korea Information Infrastructure) aims to build broadband backbone networks mainly through optical fiber cables and ATM switches and to Provide the government ministries, local authorities and non-profit organizations with high-speed, highly capable broadband access to these networks at reasonable rates. The KB service model, however, is different from other Internet Service models much in the same way SP (Service Provider) and NP (Network Provider) differ from each other. In this paper, we evaluate KII service according to various customer satisfaction indicators under the methodologies and categories as put forth in both the ACSI (American Customer Satisfaction Index) and SEM (Structural Equation Model). We use a structural equation model (SEM) to demarcate the Korea Information Infrastructure Customer Satisfaction Index (KIICSI) in relation to network service quality. The results of our study suggest some strategies for the KII Project need to be modified and effectively implemented in order to increase the satisfaction level of the KII customers.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.4 no.2
• /
• pp.95-102
• /
• 2002

Spatial interpolation has become a common procedure in converting temperature forecasts and observations at irregular points for use in regional scale ecosystem modeling and the model based decision support systems for resource management. Neglection of terrain effects in most spatial interpolations for short term temperatures may cause erroneous results in mountainous regions, where the observation network hardly covers full features of the complicated terrain. A spatial interpolation model for daytime hourly temperature was formulated based on error analysis of unsampled site with respect to the site topography. The model has a solar irradiance correction scheme in addition to the common backbone of the lapse rate - corrected inverse distance weighting. The solar irradiance scheme calculates the direct, diffuse and reflected components of shortwave radiation over any surfaces based on the sun-slope geometry and compares the sum with that over a reference surface. The deviation from the reference radiation is used to calculate the temperature correction term by an empirical conversion formula between the solar energy and the air temperature on any sloped surfaces at an hourly time scale, which can be prepared seasonally for each land cover type. When this model was applied to a 14 km by 22 km mountainous region at a 10 m horizontal resolution, the estimated hourly temperature surfaces showed a better agreement with the observed distribution than those by a conventional method.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.1B
• /
• pp.15-30
• /
• 2000

The important considerations in the design of fiber-optic networks are reliability and survivability preparing against a failure. The SDH(Synchronous Digital Hierarchy), the international standard of optical transmission, offers several network reconfiguration methods that enable network to be automatically restored from failure. One of the methods is the SHR(Self Healing Ring), which is a ring topology system. Most network providers have constructed their backbone networks with SHR architecture since it can provide survivability economically. The network architecture has eventually evolved into a multi-ring network comprised of interconnected rings. This paper addresses multi-ring network design problems is to minimize ring-construction cost. This problem can be formulated with MIP(mixed integer programming) model. However, it is difficult to solve the model within reasonable computing time on a large scale network because the model is NP-complete. Furthermore, in practice we should consider the problem of routing demands on rings to minimize total cost. This routing problem involves multiplex bundling at the intermediate nodes. A family of heuristic algorithms is presented for this problem. These algorithms include gateway selection and routing of inter-ring demands as well as load balancing on single rings. The developed heuristic algorithms are applied to some network provider's regional and long-distance transmission networks. We show an example of ring design and compare it with another ring topology design. Finally, we analysis the effect bundling.

• Journal of Internet Computing and Services
• /
• v.24 no.6
• /
• pp.91-98
• /
• 2023

Recently, according to development of artificial intelligence, a wide range of industry being automatic and optimized. Also we can find out some research of using supervised learning for deteceting defect of railway in domestic rail industry. However, there are structures other than rails on the track, and the fastener is a device that binds the rail to other structures, and periodic inspections are required to prevent safety accidents. In this paper, we present a method of reducing cost for labeling using semi-supervised and transfer model trained on rail fastener data. We use Resnet50 as the backbone network pretrained on ImageNet. At first we randomly take training data from unlabeled data and then labeled that data to train model. After predict unlabeled data by trained model, we adopted a method of adding the data with the highest probability for each class to the training data by a predetermined size. Futhermore, we also conducted some experiments to investigate the influence of the number of initially labeled data. As a result of the experiment, model reaches 92% accuracy which has a performance difference of around 5% compared to supervised learning. This is expected to improve the performance of the classifier by using relatively few labels without additional labeling processes through the proposed method.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.9
• /
• pp.34-41
• /
• 2017

Wireless mesh networks contain multi-hop routing protocols between wireless nodes and are connected to the Internet through a gateway. These networks perform a role as a backbone and are scalable for main applications. We propose the design of QoS supporting mechanisms for wireless mesh networks using software defined networking. Our proposed scheme is cost-effective and features adaptive QoS mechanisms for wireless router's resource constraints. The QoS mechanisms use software defined networking technology with Openflow protocol based on diffserv and intserv models with MPLS mechanism and RSVP respectively. A performance evaluation model is suggested to verify the validity of the proposed scheme using several QoSmetrics of the wireless mesh networks.

• Biotechnology and Bioprocess Engineering:BBE
• /
• v.10 no.3
• /
• pp.218-224
• /
• 2005

Surface modification of glutaraldehyde fixed bovine pericardium (GFBP) was success­fully carried out with hyaluronic acid (HA) derivatives. At first, HA was chemically modified with adipic dihydrazide (ADH) to introduce hydrazide functional group into the carboxyl group of HA backbone. Then, GFBP was surface modified by grafting HA-ADH to the free aldehyde groups on the tissue and the subsequent HA-ADH hydrogel coating. HA-ADH hydrogels could be prepared through selective crosslinking at low pH between hydrazide groups of HA-ADH and crosslinkers containing succinimmidyl moieties with minimized protein denaturation. When HA­ADH hydrogels were prepared at low pH of 4.8 in the presence of erythropoietin (EPO) as a model protein, EPO release was continued up to $85\%$ of total amount of loaded EPO for 4 days. To the contrary, only $30\%$ of EPO was released from HA-ADH hydrogels prepared at pH=7.4, which might be due to the denaturation of EPO during the crosslinking reaction. Because the carboxyl groups on the glucuronic acid residues are recognition sites for HA degradation by hyaluronidase, the HA-ADH hydrogels degraded more slowly than HA hydrogels prepared by the crosslinking reaction of divinyl sulfone with hydroxyl groups of HA. Following a two-week subcutaneous implantation in osteopontin-null mice, clinically significant levels of calcification were observed for the positive controls without any surface modification. However, the calcification of surface modified GFBP with HA-ADH and HA-ADH hydrogels was drastically reduced by more than $85\%$ of the positive controls. The anti-calcification effect of HA surface modification was also confirmed by microscopic analysis of explanted tissue after staining with Alizarin Red S for calcium, which followed the trend as observed with calcium quantification.