• Journal of the Korean Society of Industry Convergence
• /
• v.23 no.6_2
• /
• pp.1025-1032
• /
• 2020

In this paper, we propose a method to classify road weather conditions into rain, fog, and sun using a SVM (Support Vector Machine) classifier after extracting weather features from images acquired in real time using an optical sensor installed on a roadside post. A multi-dimensional weather feature vector consisting of factors such as image sharpeness, image entropy, Michelson contrast, MSCN (Mean Subtraction and Contrast Normalization), dark channel prior, image colorfulness, and local binary pattern as global features of weather-related images was extracted from road images, and then a road weather classifier was created by performing machine learning on 700 sun images, 2,000 rain images, and 1,000 fog images. Finally, the classification performance was tested for 140 sun images, 510 rain images, and 240 fog images. Overall classification performance is assessed to be applicable in real road services and can be enhanced further with optimization along with year-round data collection and training.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2013.04a
• /
• pp.611-616
• /
• 2013

Nowadays concrete pavement is increasing, since it is more durable than asphalt pavement. And the concrete pavement with lateral rain groove may induce high level of concrete whine noise as pure tone. In this paper, the design factor for good concrete whine noise is considered in view of tire pattern and structure design. In respect of tire pattern design, the tire having a cap tread with high center part stiffness and low shoulder part stiffness shows best concrete whine noise performance. And in respect of tire structural design, the tire with a thick center part of cap tread and low tread part stiffness show best concrete whine noise performance.

• Journal of Fashion Business
• /
• v.17 no.5
• /
• pp.120-137
• /
• 2013

Nowadays, due to the increase of the aging population and the decreased size of the traditional nuclear family, pet dogs are considered to be important member of the family. The pet industry has been growing every year; subsequently, is essential to develop practical patterns and designs which are appropriate for the structure and movement of pet dogs. However, studies for developing a new form of pet dog's wear patterns and designs are limited. This study develops an all-in-one pattern which is based on pet dog's size, designs pet dog's wear items based on all-in-one patterns, and provides support to the manufacture of pet dog's wear. The results are as follow: 1) An analysis for the categories and design characteristics of all-in-one pet dog's wear according to the research from the online pet dog's wear shopping sites is provided. Eight categories are classified as follows: rompers, body suit, coveralls, training suit, special clothing, overalls, rain coat, and hanbok. 2) The production of all-in-one pet dog's wear basic pattern based on the basic bodice pattern from the prior studies and the manufacturer's patterns are being displayed. 3) Developments of the patterns and designs for eight all-in-one clothing consider the design characteristics and the situations. 4) Developments of the previous eight items illustrate the results of this research.

• KCID journal
• /
• v.14 no.1
• /
• pp.57-66
• /
• 2007

Temporal changes in total nitrogen (T-N) and phosphorus (T-P) concentrations in paddy floodwater in response to fertilization under rain-shielding pot and small-scaled field conditions were investigated. On the basis of the changing patterns, suggestions for the use of fertilization factors, such as days after fertilization, in developing models for the estimation of T-N and T-P loads from paddy fields were made. Total N concentration was susceptible to fertilization, showing a peak concentration right after fertilization followed by a decreasing pattern with the elapse of days after fertilization. The decreasing pattern of T-N concentration followed the first- order kinetics, indicating that the models are likely to be an exponential equation using days after fertilization as an independent variable. Comparison between the pot and field experiments conducted with soils different in soil fertility revealed that indigenous soil N concentration significantly affected T-N concentration, and this suggests that soil N status can be used as the second variable for the models. Meanwhile, temporal changes in T-P concentration did not respond to P fertilization as sensitively as T-N. In combination with other published results, our study suggests that rainfall intensity and other factors associated with farming activities that are likely to cause disturbance of soil particles containing P may be used as possible variables for the models.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.35 no.6
• /
• pp.1219-1228
• /
• 2015

The Fixed Area ARFs (Area Reduction Factors) method has limitations in providing exact information about spatial distribution due to the lack of enough density of rain gauge stations. In this study the storm-centered ARF was evaluated between frontal and typhoon storm events utilizing radar precipitation. In estimating storm-centered ARFs, in order to consider the horizontal advection, direction, and spatial distribution of rain cells, the rotational angle of rainfall of each rainfall event and the optimum areal rainfall within the spatial rain cell envelope was taken into account. Compared with the frontal storm, the ARF of typhoon storm shows narrow range of variability. It is noted that the ARFs of frontal storm increases with the rainfall duration, but those of typhoon storm shows opposite pattern. As a result the typhoon ARFs appear greater than frontal ARFs for 1~3 hours of duration, but less for more than 6 hours of duration.

• Journal of Environmental Science International
• /
• v.26 no.3
• /
• pp.289-302
• /
• 2017

During the research period, error analysis of the amount of daily precipitation was performed with data obtained from 2DVD, Parsivel, and AWS, and from the results, 79 days were selected as research days. According to the results of a synoptic meteorological analysis, these days were classified into 'LP type, CF type, HE type, and TY type'. The dates showing the maximum daily precipitation amount and precipitation intensity were 'HE type and CF type', which were found to be attributed to atmospheric instability causing strong ascending flow, and leading to strong precipitation events. Of the 79 days, most days were found to be of the LP type. On July 27, 2011 the daily precipitation amount in the Korean Peninsula reached over 80 mm (HE type). The leading edge of the Northern Pacific high pressure was located over the Korean Peninsula with unstable atmospheric conditions and inflow of air with high temperature and high humidity caused ascending flow, 120 mm/h with an average precipitation intensity of over 9.57 mm/h. Considering these characteristics, precipitation in these sample dates could be classified into the convective rain type. The results of a precipitation scale distribution analysis showed that most precipitation were between 0.4-5.0 mm, and 'Rain' size precipitation was observed in most areas. On July 9, 2011, the daily precipitation amount was recorded to be over 80 mm (CF type) at the rainy season front (Jangma front) spreading across the middle Korean Peninsular. Inflow of air with high temperature and high humidity created unstable atmospheric conditions under which strong ascending air currents formed and led to convective rain type precipitation.

• Journal of Wetlands Research
• /
• v.13 no.1
• /
• pp.35-44
• /
• 2011

In order to explore vegetation adaptation to climate variability and the impacts on water balance dynamics, the inter-regional and the inter-annual variability of both water availability and vegetation productivity are investigated. The Horton index, which is the ratio between actual evapotranspiration and catchment wetting as a measure of vegetation water use at catchment-scale, is revisited to quantify the effects of growing-season water availability on hydrologic partitioning at catchment scale. It is shown that the estimated Horton index is relatively constant irrespective of inter-annual climate variability. In addition, the Horton index is compared with catchment-scale vegetation rain use efficiency. The results show that there is an interesting pattern in the response of vegetation water use to water availability. When water becomes the limiting factor for vegetation productivity, the catchment-scale vegetation rain use efficiency converges to a common maximum value in agreement with earlier findings at the ecosystem level.

• Atmosphere
• /
• v.26 no.2
• /
• pp.321-336
• /
• 2016

Damages caused by torrential rain occur every year in Korea and summer time convection can cause strong thunderstorms to develop which bring dangerous weather such as torrential rain, gusts, and flash flooding. On 6 August 2013 a sudden torrential rain concentrated over the inland of Southern Korean Peninsula occurred. This was an event characterized as a mesoscale multicellular convection. The purpose of this study is to analyze the conditions of the multicellular convection and the synoptic and mesoscale nature of the system development. To this end, dynamical and thermodynamic analyses of surface and upper-level weather charts, satellite images, soundings, reanalysis data and WRF model simulations are performed. At the beginning stage there was a cool, dry air intrusion in the upper-level of the Korean Peninsula, and a warm humid air flow from the southwest in the lower-level creating atmospheric instability. This produced a single cell cumulonimbus cloud in the vicinity of Baengnyeongdo, and due to baroclinic instability, shear and cyclonic vorticity the cloud further developed into a multicellular convection. The cloud system moved southeast towards Seoul metropolitan area accompanied by lightning, heavy precipitation and strong wind gusts. In addition, atmospheric instability due to daytime insolation caused new convective cells to develop in the upstream part of the Sobaek Mountain which merged with existing multicellular convection creating a larger system. This case was unusual because the system was affected little by the upper-level jet stream which is typical in Korea. The development and propagation of the multicellular convection showed strong mesoscale characteristics and was not governed by large synoptic-scale dynamics. In particular, the system moved southeast crossing the Peninsula diagonally from northwest to southeast and did not follow the upper-level westerly pattern. The analysis result shows that the movement of the system can be determined by the vertical wind shear.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.55 no.6
• /
• pp.19-30
• /
• 2013

This study is to evaluate the use of dual-polarization radar data for storm runoff modeling in Namgang dam (2,293 $km^2$) watershed using KIMSTORM (Grid-based KIneMatic wave STOrm Runoff Model). The Bisl dual-polarization radar data for 3 typhoons (Khanun, Bolaven, Sanba) and 1 heavy rain event in 2012 were obtained from Han River Flood Control Office. Even the radar data were overall less than the ground data in areal average, the spatio-temporal pattern between the two data was good showing the coefficient of determination ($R^2$) and bias with 0.97 and 0.84 respectively. For the case of heavy rain, the radar data caught the rain passing through the ground stations. The KIMSTORM was set to $500{\times}500$ m resolution and a total of 21,372 cells (156 rows${\times}$137 columns) for the watershed. Using 28 ground rainfall data, the model was calibrated using discharge data at 5 stations with $R^2$, Nash and Sutcliffe Model Efficiency (ME) and Volume Conservation Index (VCI) with 0.85, 0.78 and 1.09 respectively. The calibration results by radar rainfall showed $R^2$, ME and VCI were 0.85, 0.79, and 1.04 respectively. The VCI by radar data was enhanced by 5 %.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.18 no.3
• /
• pp.297-305
• /
• 2008

Recently, kernel methods have attracted great interests in the areas of pattern classification, function approximation, and anomaly detection. The role of the kernel is particularly important in the methods such as SVM(support vector machine) and KPCA(kernel principal component analysis), for it can generalize the conventional linear machines to be capable of efficiently handling nonlinearities. This paper considers the problem of combining radar and rain gauge observations utilizing the regression approach based on the kernel-based gaussian process and support vector learning. The data-assimilation results of the considered methods are reported for the radar and rain gauge observations collected over the region covering parts of Gangwon, Kyungbuk, and Chungbuk provinces of Korea, along with performance comparison.