• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.6 no.12
• /
• pp.3315-3337
• /
• 2012

Crowded scenes analysis is a challenging topic in computer vision field. How to detect diverse motion patterns in crowded scenarios from videos is the critical yet hard part of this problem. In this paper, we propose a novel approach to mining motion patterns by utilizing motion information during both long-term period and short interval simultaneously. To capture long-term motions effectively, we introduce Motion History Image (MHI) representation to access to the global perspective about the crowd motion. The combination of MHI and optical flow, which is used to get instant motion information, gives rise to discriminative spatial-temporal motion features. Benefitting from the robustness and efficiency of the novel motion representation, the following motion pattern mining is implemented in a completely unsupervised way. The motion vectors are clustered hierarchically through automatic hierarchical clustering algorithm building on the basis of graphic model. This method overcomes the instability of optical flow in dealing with time continuity in crowded scenes. The results of clustering reveal the situations of motion pattern distribution in current crowded videos. To validate the performance of the proposed approach, we conduct experimental evaluations on some challenging videos including vehicles and pedestrians. The reliable detection results demonstrate the effectiveness of our approach.

• KIPS Transactions on Software and Data Engineering
• /
• v.2 no.3
• /
• pp.173-186
• /
• 2013

Web-based map services provide various geographic information that users want to get by continuous updating of data. Those map services provide different information for a geographic object respectively. It causes several problems, and most of all various information cannot be integrated and provided. To resolve the problem, this paper proposes a system which can integrate diverse geographic information and provide users rich geographic information. In this paper, a hybrid schema matching (HSM) algorithm is proposed and the algorithm is a mixture of the adapter-based semantic processing method, static semantic management-based approach, and dynamic semantic management-based approach. A comparative evaluation is described to show effectiveness of the proposed algorithm. The proposed algorithm in this paper improves the accuracy of schema matching because of registration and management of schemas of new semantic information. The proposal enables vocabulary-based schema matching using various schemas, and it thus also supports high usability. Finally, the proposed algorithm is cost-effective by providing the progressive extension of relationships between schema meanings.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.11
• /
• pp.2047-2058
• /
• 2007

The business environment of enterprises could be difficult to obviate redundancy to filtrate data source occurred on data integrated to standard rules and meta-data and to produce integration of data and single viewer in geographical and spatial distributed environment. Specially, To can interchange various data from a heterogeneous system or various applications without types and forms and synchronize continually exactly integrated information#s is of paramount concern. Therefore data hub system based on SQL/XMDR message to overcome a problem of meaning interoperability occurred on exchanging or jointing between each legacy systems are proposed in this paper. This system use message mapping technique of query transform system to maintain data modified in real-time on cooperating data. It can consistently maintain data modified in realtime on exchanging or jointing data for cooperating legacy systems, it improve clarity and availability of data by providing a single interface on data retrieval.

• Korean Journal of Remote Sensing
• /
• v.11 no.2
• /
• pp.1-16
• /
• 1995

This study is conducted to eveluate the potential of a GIS to assist an application problem. GIS has been applied to rainfall-runoff modeling over Soyang area. Various rainfall-runoff models have been developed over the years. A distributed rainfall-runoff model is selected because it considers the topographic characteristics over the basin. GIS can handle the spatial data to enhance the modeling. GRASS-a public domain GIS S/W-is used for GIS tools. Digital database is generated, including soil map, vegetation map, digital elevation model, basin and subbasin map, and water stream. The inpu data for the model has been generated and manupulated using GIS. The database, model and GIS are integrated for on-line operation. The inflow hydrographs are tested for the flood of Sept., 1990. This shows the promising results even without the calibration.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2018.05a
• /
• pp.148-148
• /
• 2018

Land subsidence caused by excessive groundwater pumping is a serious hydro-geological hazard. The spatial variability in land use, unbalanced groundwater extraction and aquifer characteristics are the key factors which make the problem more difficult to monitor using conventional methods. This study uses the European Space Agency (ESA) Sentinel satellites to investigate and monitor land subsidence varying with different land covers and groundwater use in the arid Quetta valley, Pakistan. The Persistent Scattering Differential Interferometry of Synthetic Aperture Radar (PS-DInSAR) method was used to develop 28 subsidence interferograms of the study area for the period between 16 Oct 2014 and 06 Oct 2016 using ESA's Sentinel-1 SAR data. The uncertainty of DInSAR result is first minimized by removing the dynamic effect caused by atmospheric factors and then filtered using the radar Amplitude Dispersion Index (ADI) to select only the stable pixels. Finally the subsidence maps were generated by spatially interpolating the land subsidence at the stable pixels, the comparison of DInSAR subsidence with GPS readings showed an R 2 of 0.94 and mean absolute error of $5.7{\pm}4.1mm$. The subsidence maps were also analysed for the effect of aquifer type and 4 land covers which were derived from Sentienl-2 multispectral images. The analysis show that during the two year period, the study area experienced highly non-linear land subsidence ranging from 10 to 280 mm. The subsidence at different land covers was significantly different from each other except between the urban and barren land. The barren land and seasonally cultivated area show minor to moderate subsidence while the orchard and urban area with high groundwater extraction rate showed excessive amount of land subsidence. Moreover, the land subsidence and groundwater drawdown was found to be linearly proportional to each other.

• Korean Journal of Remote Sensing
• /
• v.37 no.5_1
• /
• pp.1061-1070
• /
• 2021

In this manuscript, we tried to improve the performance of the FC-DenseNet by applying an attention gate for the classification of cropping areas. The attention gate module could facilitate the learning of a deep learning model and improve the performance of the model by injecting of spatial/spectral weights to each feature map. Crop classification was performed in the onion and garlic regions using a proposed deep learning model in which an attention gate was added to the skip connection part of FC-DenseNet. Training data was produced using various PlanetScope satellite imagery, and preprocessing was applied to minimize the problem of imbalanced training dataset. As a result of the crop classification, it was verified that the proposed deep learning model can more effectively classify the onion and garlic regions than existing FC-DenseNet algorithm.

• Journal of Agricultural Extension & Community Development
• /
• v.27 no.4
• /
• pp.173-184
• /
• 2020

Is the Fourth Industrial Revolution a revolution for cities only? Through the Fourth Industrial Revolution, Korea has entered quickly in the influence area of intelligent information technology such as IoT, AI, Big data, Cloud, ICT, Digital twin. However, as the information gap between the rural zone and the urban zone worsens, a policy was needed to reduce such a gap. Therefore, this research analyzed EU's smart village project, and investigated the problem and improvement of the actual smart village through the interview and field study with the person in charge of the actual smart village project in Korea. Based on the analytic result, 5 plans were deduced to improve Korea's smart village project. First, make the realistic adjustment of project period to assure the sustainability of smart village; second, make the new establishment of the department in charge of smart village project; third, construct the system of integrating and cooperating the policy that can embrace all the rural zone and the urban zone; the fourth, expand the application area of customized ICT technology according to the new rural policy environment; and finally introduce the residents' capacity development project through the rural guidance project.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.13 no.6
• /
• pp.1353-1364
• /
• 2018

This study proposed a method to detect Cochlodinium polykrikoides red tide pixels in satellite images using a logistic regression model of machine learning technique under Imbalanced data. The spectral profiles extracted from red tide, clear water, and turbid water were used as training dataset. 70% of the entire data set was extracted and used for as model training, and the classification accuracy of the model was evaluated using the remaining 30%. At this time, the white noise was added to the spectral profile of the red tide, which has a relatively small number of data compared to the clear water and the turbid water, and over-sampling was performed to solve the unbalanced data problem. As a result of the accuracy evaluation, the proposed algorithm showed about 94% classification accuracy.

• Journal of Ecology and Environment
• /
• v.42 no.4
• /
• pp.308-316
• /
• 2018

Background: Genus Sasa, dwarf bamboos, are considered to be species that lower biodiversity in the temperate forests of East Asia. Although they have a long interval, they, the monocarpic species, have a unique characteristic of large-scale synchronized flowering. Therefore, once they have flowered and then declined, it may be an opportunity for suppressed surrounding species. A previous study reported that Sasa borealis showed specialized flowering nationwide with a peak in 2015. However, this was based on data from a social network service and field survey at Mt. Jeombong. Therefore, we investigated S. borealis in the forests of five national parks in order to determine whether this rare synchronized flowering occurred nationwide, as well as its spatial distribution. Results: We found a total of 436 patches under the closed canopy of Quercus mongolica-dominated deciduous forests in the surveyed transects from the five national parks. Of these patches, 75% occupied a whole slope area, resulting in an enormous area. The patch area tended to be larger in the southern parks. Half (219 patches) of the patches flowered massively. Among them, 76% bloomed in 2015, which was consistent with the results of the previous report. The flowering rate varied from park to park with that of Mt. Seorak being the highest. The culms of the flowering patches were significantly taller (F = 93.640, p < 0.000) and thicker (F = 61.172, p < 0.000). Following the event, the culms of the flowering patches declined, providing a good opportunity for the suppressed plant species. The concurrent massive flowering of the mature patches was believed to be triggered by some stress such as a spring drought. Conclusion: We confirmed that the rare synchronized flowering of S. borealis occurred with a peak in 2015 nationwide. In addition, we explored that S. borealis not only monopolized an enormous area, but also dominated the floors of the late-successional Q. mongolica-dominated deciduous forests. This presents a major problem for Korean forests. As it declined simultaneously after flowering, there are both possibilities of forest regeneration or resettlement of S. borealis by massively produced seeds.

• Journal of ILASS-Korea
• /
• v.23 no.4
• /
• pp.192-204
• /
• 2018

In this study, a 3D CFD analysis method for the combustion process was established for a low calorific value syngas-diesel dual-fuel engine operating under very lean fuel-air mixture condition. Also, the accuracy of computational analysis was evaluated by comparing the experimental results with the computed ones. To simulate the combustion for the dual-fuel engine, a new dual-fuel chemical kinetics set was used that was constituted by merging two verified chemical kinetic sets: n-heptane (173 species) for diesel and Gri-mech 3.0 (53 species) for syngas. For dual-fuel mode operations, the early stage of combustion was dominated by the fuel burning inside or near the spray plume. After which, the flame propagated into the syngas in the piston bowl and then proceeded toward the syngas in the squish zone. With the baseline injection system and piston shape, a significant amount of unburned syngas was discharged. To solve this problem, effects of the injection parameters and piston shape on combustion characteristics were analyzed by calculation. The change in injection variables toward increasing the spray plume volume or the penetration length were effective to cause fast burning in the vicinity of TDC by widening the spatial distribution of diesel acting as a seed of auto-ignition. As a result, the unburned syngas fraction was reduced. Changing the piston shape with the shallow depth of the piston bowl and 20% squish area ratio had a significant effect on the combustion pattern and lessened the unburned syngas fraction by half.