• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.23 no.3
• /
• pp.145-151
• /
• 2023

3D-CNN is one of the deep learning techniques for learning time series data. Such three-dimensional learning can generate many parameters, so that high-performance machine learning is required or can have a large impact on the learning rate. When learning dynamic hand-gestures in spatiotemporal domain, it is necessary for the improvement of the efficiency of dynamic hand-gesture learning with 3D-CNN to find the optimal conditions of input video data by analyzing the learning accuracy according to the spatiotemporal change of input video data without structural change of the 3D-CNN model. First, the time ratio between dynamic hand-gesture actions is adjusted by setting the learning interval of image frames in the dynamic hand-gesture video data. Second, through 2D cross-correlation analysis between classes, similarity between image frames of input video data is measured and normalized to obtain an average value between frames and analyze learning accuracy. Based on this analysis, this work proposed two methods to effectively select input video data for 3D-CNN deep learning of dynamic hand-gestures. Experimental results showed that the learning interval of image data frames and the similarity of image frames between classes can affect the accuracy of the learning model.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2007.10a
• /
• pp.578-581
• /
• 2007

For tracing tag locations, a trajectories should be modeled and indexed in radio frequency identification (RFID) systems. The trajectory of a tag can be represented as a line that connects two spatiotemporal locations captured when the tag enters and leaves the vicinity of a reader. If a tag enters but does not leave a reader, its trajectory is represented only as a point captured at entry. Because the information that the tag stays in the reader is missing from the trajectory represented only as a point, we should extend the region of a query to find the tag that remains in a reader. In this paper, we propose an interval data model of tag's trajectory in order to solve the problem. Trajectories of tags are represented as two kinds of intervals; dynamic intervals which are time-dependent lines and static intervals which are fixed lines. We also show that the interval data model has better performance than others with a cost model

• Journal of Korean Society on Water Environment
• /
• v.30 no.5
• /
• pp.549-558
• /
• 2014

Occurrence frequency of flood and drought tends to increase in last a few decades, leading to social and economic damage since the abnormality of climate changes is one of the causes for hydrologic facilities by exceedance its designed tolerance. Soil and Water Assessment Tool (SWAT) model was used in the study to estimate temporal variance of groundwater recharge and baseflow. It was limited to consider recession curve coefficients in SWAT model calibration process, thus the recession curve coefficient was estimated by the Baseflow Filter Program (BFLOW) before SWAT model calibration. Precipitation data were estimated for 2014 to 2100 using three models which are GFDL-ESM2G, IPSL-CM5A-LR, and MIROC-ESM with Representative Concentration Pathways (RCP) scenario. SWAT model was calibrated for the Soyang watershed with NSE of 0.83, and $R^2$ of 0.89. The percentage to precipitation of groundwater recharge and baseflow were 27.6% and 17.1% respectively in 2009. Streamflow, groundwater recharge, and baseflow were estimated to be increased with the estimated precipitation data. GFDL-ESM2g model provided the most large precipitation data in the 2025s, and IPSL-CM5A-LR provided the most large precipitation data in the 2055s and 2085s. Overall, groundwater recharge and baseflow displayed similar trend to the estimated precipitation data.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.39 no.6
• /
• pp.49-58
• /
• 2002

Based on electrophysiological retina mechanism, a retina model is proposed, which has similar response characteristics compared with the real primate retina. Photoreceptors, horizontal cells, and bipolar cells are modeled based on the previously studied retina models. And amacrine cells known to have relation to movements detection, and bipolar cell terminals are newly modeled using 3 NDP mechanism. The proposed model verified by analyzing the spatial response characteristics to stationary and moving stimuli, and characteristics for different speeds. Through this retina model, human vision system could be applied to computer vision systems for movement detection, and it could be the basic research for the implantable artificial retina.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.32 no.4_1
• /
• pp.335-341
• /
• 2014

The main goal of this study is on developing a proper brand-new data of national base map and Data Based(DB) model for new information technology environments. To achieve this goal, we generated a brand-new Hybrid spatial information model which is specialized in the spatio-temporal map structure, the framework map for information integration, and the multiple-layered topology structure. The DB structure was designed to reflect the change of objections by adding a new dimension of 'time' in the spartial information, while the infrastructure was able to connect/converge with other information by giving the unique ID and multi-scale fusion map structure. Furthermore, the topology and multi visualization structure, including indoor and basement information, were designed to overcome limitations of expressing in 2 dimension map. The result from the performance test, which was based on the Hybrid spatial information model, confirms the possibility in advanced national base map and conducted DB model through implementing various information and spatiotemporal connections.

• KCID journal
• /
• v.14 no.1
• /
• pp.50-56
• /
• 2007

In this study, the Soil and Water Assessment Tool (SWAT) model was used to assess spatiotemporal effects on watershed runoff and sediment characteristics due to land uses changes from 1999 to 2002 at the small watershed, located in Chuncheon-si, Gangwon province. The annual average flow rate of Scenario I (long-term simulation using land use of 1990), II (long-term simulation using land use of 1996), III(long-term simulation using land use of 200) and IV(simulation using land use of 1990, 1995, and 2000) in long-term simulation) using the SWAT model were 29,997,043 m3, 29,992,628 m3, 29,811,191 m3 and 29,931,238 m3, respectively. It was shown that there was no significant changes in estimated flow rate because no significant changes in land uses between 1990 and 2000 were observed. The annual average sediment loads of Scenarios I, II, III and IV for 15 year period were 36,643 kg/ha, 45,340 kg/ha , 27,195 kg/ha and 35,545 kg/ha, respectively. The estimated annual sediment loads from Scenarios I, II, and III, were different from that from the scenario IV, considering spatio-temporal changes in land use and meterological changes over the years, by 10%, 127%, and temporal changes in land use and meterological changes over the years, by 10%, 127%, and 77%. This can be explained in land use changes in high soil erosion potential areas, such as upland areas, within the study watershed. The comparison indicates that changes in land uses upland areas, within the study watershed. The comparison indicates that changes in land uses can affect on sediment yields by more than 10%, which could exceed the safety factor of 10% in Total Maximum Daily Loads (TMDLs). It is, therefore, recommended that not only the temporal analysis with the weather input data but also spatial one with different land uses need to be considered in long-term hydrology and sediment simulating using the SWAT model

• Journal of Soil and Groundwater Environment
• /
• v.26 no.6
• /
• pp.165-175
• /
• 2021

Groundwater is used in many areas in food industry such as food manufacturing, food processing, cooking, and liquor industry etc. in Korea. As groundwater occupies a large portion of food industry, it is necessary to predict deterioration of water quality to ensure the safety of food water since using undrinkable groundwater has a ripple effect that can cause great harm or anxiety to food users. In this study, spatiotemporal data aggregation method was used in order to obtain spatially representative data, which enable prediction of groundwater quality change in a small watershed. In addition, a highly reliable predictive model was developed to estimate long-term changes in groundwater quality by applying a non-parametric segmented regression technique. Two pilot watersheds were selected where a large number of companies use groundwater for food water, and the appropriateness of the model was assessed by comparing the model-produced values with those obtained by actual measurements. The result of this study can contribute to establishing a customized food water management system utilizing big data that respond quickly, accurately, and preemptively to changes in groundwater quality and pollution. It is also expected to contribute to the improvement of food safety management.

• The Korean Journal of Applied Statistics
• /
• v.35 no.5
• /
• pp.603-617
• /
• 2022

Following the September 11, 2001 terrorist attacks, the United States declared war on terror and invaded Afghanistan and Iraq, winning quickly. However, interest in analyzing terrorist activities has developed as a result of a significant amount of time being spent on the post-war stabilization effort, which failed to minimize the number of terrorist activities that occurred later. Based on terrorist data from 2003 to 2010, this study utilized a Bayesian hierarchical model to forecast the terrorist threat in 2011. The model depicts spatiotemporal dependence with predictors such as population and religion by autonomous district. The military commander in charge of the region can utilize the forecast value based on the our model to prevent terrorism by deploying forces efficiently.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2017.05a
• /
• pp.260-260
• /
• 2017

This study aims to compare the performance of TRIGRS (Transient Rainfall Infiltration and Grid-based Regional Slope-stability model) and TiVaSS (Time-variant Slope Stability model) in the prediction of rainfall-induced shallow landslides. TRIGRS employs one-dimensional (1-D) subsurface flow to simulate the infiltration rate, whereas a three-dimensional (3-D) model is utilized in TiVaSS. The former has been widely used in landslide modeling, while the latter was developed only recently. Both programs are used for the spatiotemporal prediction of shallow landslides caused by rainfall. The present study uses the July 2011 landslide event that occurred in Mt. Umyeon, Seoul, Korea, for validation. The performance of the two programs is evaluated by comparison with data of the actual landslides in both location and timing by using a landslide ratio for each factor of safety class ( index), which was developed for addressing point-like landslide locations. In addition, the influence of surface flow on landslide initiation is assessed. The results show that the shallow landslides predicted by the two models have characteristics that are highly consistent with those of the observed sliding sites, although the performance of TiVaSS is slightly better. Overland flow affects the buildup of the pressure head and reduces the slope stability, although this influence was not significant in this case. A slight increase in the predicted unstable area from 19.30% to 19.93% was recorded when the overland flow was considered. It is concluded that both models are suitable for application in the study area. However, although it is a well-established model requiring less input data and shorter run times, TRIGRS produces less accurate results.

• Korean Journal of Remote Sensing
• /
• v.39 no.5_3
• /
• pp.933-948
• /
• 2023

Atmospheric aerosols not only have adverse effects on human health but also exert direct and indirect impacts on the climate system. Consequently, it is imperative to comprehend the characteristics and spatiotemporal distribution of aerosols. Numerous research endeavors have been undertaken to monitor aerosols, predominantly through the retrieval of aerosol optical depth (AOD) via satellite-based observations. Nonetheless, this approach primarily relies on a look-up table-based inversion algorithm, characterized by computationally intensive operations and associated uncertainties. In this study, a novel high-resolution AOD direct retrieval algorithm, leveraging machine learning, was developed using top-of-atmosphere reflectance data derived from the Geostationary Ocean Color Imager-II (GOCI-II), in conjunction with their differences from the past 30-day minimum reflectance, and meteorological variables from numerical models. The Light Gradient Boosting Machine (LGBM) technique was harnessed, and the resultant estimates underwent rigorous validation encompassing random, temporal, and spatial N-fold cross-validation (CV) using ground-based observation data from Aerosol Robotic Network (AERONET) AOD. The three CV results consistently demonstrated robust performance, yielding R²=0.70-0.80, RMSE=0.08-0.09, and within the expected error (EE) of 75.2-85.1%. The Shapley Additive exPlanations(SHAP) analysis confirmed the substantial influence of reflectance-related variables on AOD estimation. A comprehensive examination of the spatiotemporal distribution of AOD in Seoul and Ulsan revealed that the developed LGBM model yielded results that are in close concordance with AERONET AOD over time, thereby confirming its suitability for AOD retrieval at high spatiotemporal resolution (i.e., hourly, 250 m). Furthermore, upon comparing data coverage, it was ascertained that the LGBM model enhanced data retrieval frequency by approximately 8.8% in comparison to the GOCI-II L2 AOD products, ameliorating issues associated with excessive masking over very illuminated surfaces that are often encountered in physics-based AOD retrieval processes.