• Korean Journal of Remote Sensing
• /
• v.37 no.5_1
• /
• pp.871-884
• /
• 2021

The purpose of this study was to determine ways to increase efficiency in constructing and verifying artificial intelligence learning data on land cover using aerial and satellite images, and in applying the data to AI learning algorithms. To this end, multi-resolution datasets of 0.51 m and 10 m each for 8 categories of land cover were constructed using high-resolution aerial images and satellite images obtained from Sentinel-2 satellites. Furthermore, fine data (a total of 17,000 pieces) and coarse data (a total of 33,000 pieces) were simultaneously constructed to achieve the following two goals: precise detection of land cover changes and the establishment of large-scale learning datasets. To secure the accuracy of the learning data, the verification was performed in three steps, which included data refining, annotation, and sampling. The learning data that wasfinally verified was applied to the semantic segmentation algorithms U-Net and DeeplabV3+, and the results were analyzed. Based on the analysis, the average accuracy for land cover based on aerial imagery was 77.8% for U-Net and 76.3% for Deeplab V3+, while for land cover based on satellite imagery it was 91.4% for U-Net and 85.8% for Deeplab V3+. The artificial intelligence learning datasets on land cover constructed using high-resolution aerial and satellite images in this study can be used as reference data to help classify land cover and identify relevant changes. Therefore, it is expected that this study's findings can be used in the future in various fields of artificial intelligence studying land cover in constructing an artificial intelligence learning dataset on land cover of the whole of Korea.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.24 no.3
• /
• pp.83-98
• /
• 2021

Urban green space is an important component for enhancing urban ecosystem health. Thus, identifying the spatial structure of urban green space is required to manage a healthy urban ecosystem. The Ministry of Environment has provided the level 3 land cover map(the highest (1m) spatial resolution map) with a total of 41 classes since 2010. However, specific urban green information such as street trees was identified just as grassland or even not classified them as a vegetated area in the map. Therefore, this study classified detailed urban green information(i.e., tree, shrub, and grass), not included in the existing level 3 land cover map, using two types of high-resolution(<1m) remote sensing data(i.e., airborne LiDAR and RGB ortho imagery) in Suwon, South Korea. U-Net, one of image segmentation deep learning approaches, was adopted to classify detailed urban green space. A total of three classification models(i.e., LRGB10, LRGB5, and RGB5) were proposed depending on the target number of classes and the types of input data. The average overall accuracies for test sites were 83.40% (LRGB10), 89.44%(LRGB5), and 74.76%(RGB5). Among three models, LRGB5, which uses both airborne LiDAR and RGB ortho imagery with 5 target classes(i.e., tree, shrub, grass, building, and the others), resulted in the best performance. The area ratio of total urban green space(based on trees, shrub, and grass information) for the entire Suwon was 45.61%(LRGB10), 43.47%(LRGB5), and 44.22%(RGB5). All models were able to provide additional 13.40% of urban tree information on average when compared to the existing level 3 land cover map. Moreover, these urban green classification results are expected to be utilized in various urban green studies or decision making processes, as it provides detailed information on urban green space.

• Journal of IKEEE
• /
• v.26 no.3
• /
• pp.408-415
• /
• 2022

In this paper, we propose a 3D point cloud reconstruction technique from 2D images using efficient feature map extraction network. The originality of the method proposed in this paper is as follows. First, we use a new feature map extraction network that is about 27% efficient than existing techniques in terms of memory. The proposed network does not reduce the size to the middle of the deep learning network, so important information required for 3D point cloud reconstruction is not lost. We solved the memory increase problem caused by the non-reduced image size by reducing the number of channels and by efficiently configuring the deep learning network to be shallow. Second, by preserving the high-resolution features of the 2D image, the accuracy can be further improved than that of the conventional technique. The feature map extracted from the non-reduced image contains more detailed information than the existing method, which can further improve the reconstruction accuracy of the 3D point cloud. Third, we use a divergence loss that does not require shooting information. The fact that not only the 2D image but also the shooting angle is required for learning, the dataset must contain detailed information and it is a disadvantage that makes it difficult to construct the dataset. In this paper, the accuracy of the reconstruction of the 3D point cloud can be increased by increasing the diversity of information through randomness without additional shooting information. In order to objectively evaluate the performance of the proposed method, using the ShapeNet dataset and using the same method as in the comparative papers, the CD value of the method proposed in this paper is 5.87, the EMD value is 5.81, and the FLOPs value is 2.9G. It was calculated. On the other hand, the lower the CD and EMD values, the better the accuracy of the reconstructed 3D point cloud approaches the original. In addition, the lower the number of FLOPs, the less memory is required for the deep learning network. Therefore, the CD, EMD, and FLOPs performance evaluation results of the proposed method showed about 27% improvement in memory and 6.3% in terms of accuracy compared to the methods in other papers, demonstrating objective performance.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2005.05b
• /
• pp.1011-1015
• /
• 2005

본 실험에서는 부력 조건이 달라질 때 단공방류구에서 정체수역으로 수평방류되는 부력제트의 거동을 규명하였다. LIF (Laser Induced Fluorescence) 시스템을 이용하여 수행하였는데, LIF 시스템은 고해상도의 이미지를 취득할 수 있어 데이터의 정확도가 높으며, 동시에 한 평면상의 농도장을 일시에 측정할 수 있는 장점이 있는 기술이다. LIF 시스템은 크게 세부분으로 구성되어 있는데 방출시스템, 포착시스템, 처리시스템이 그것이다. 실험 조건을 고려해서 온수를 이용하여 주변수와의 밀도차를 재현하였으며, LIF 시스템의 추적입자로 형광염료 Rhodamine B를 사용하였다. 또한, 실험 데이터 취득과정에서 필요한 검정과정을 수행하였는데, LIF 시스템에서 검정과정은 레이저 입사광의 강도가 불균등한 분포를 가지는 점과 주변수의 매질에 의한 근의 감쇠가 발생하는 문제를 해결하기 위한 것이다. LIF 시스템은 부력제트의 농도장을 매우 정밀하게 측정할 수 있는데, 방류밀도 Froude 수가 변함에 따라 측정된 순간이미지를 통해 제트의 진화과정을 상세하고 가시적으로 확인할 수 있었다. 검정과정을 거친 농도 종단면에서 중심선의 연장선이 LIF 시스템에 의해 측정된 순간이미지의 중심선 궤적과 거의 일치하는 것도 알 수 있었다. 또한 LIF 시스템을 통해서 취득된 단일수평부력제트의 궤적과 중심선 희석률을 기존의 상용모형인 VISJET과 CORMIX1에 의해 예측된 결과와 비교$\cdot$분석한 결과, 제트 중심선 궤적의 경우, LIF 시스템을 이용한 측정값은 대체로 VISJET 모형의 결과와 일치하는 것으로 밝혀졌다. 중심선 희석률의 경우, LIF 측정값은 대체로 CORMIX1 모형, Cederwall(1968)의 경험식과 일치하는 경향을 보였다.0\%$일 때가 밸브를 $60\%$와 $80\%$ 개폐시켰을 때보다 $0.3kg/cm^2,\;0.29kg/cm^2$ 낮게 나타나 밸브를 전체 개방 했을 때 관로내의 수압이 상수설계기준에 적합한 수압을 유지함을 알 수 있다. 상수관로 설계 기준에서는 관로내 수압을 $1.5\~4.0kg/cm^2$으로 나타내고 있는데 $6kg/cm^2$보다 과수압을 나타내는 경우가 $100\%$로 밸브를 개방하였을 때보다 $60\%,\;80\%$ 개방하였을 때가 더 빈번히 발생하고 있으므로 대상지역의 밸브 개폐는 $100\%$ 개방하는 것이 선계기준에 적합한 것으로 나타났다. 밸브 개폐에 따른 수압 변화를 모의한 결과 밸브 개폐도를 적절히 유지하여 필요수량의 확보 및 누수방지대책에 활용할 수 있을 것으로 판단된다.8R(mm)(r^2=0.84)$로 지수적으로 증가하는 경향을 나타내었다. 유거수량은 토성별로 양토를 1.0으로 기준할 때 사양토가 0.86으로 가장 작았고, 식양토 1.09, 식토 1.15로 평가되어 침투수에 비해 토성별 차이가 크게 나타났다. 이는 토성이 세립질일 수록 유거수의 저항이 작기 때문으로 생각된다. 경사에 따라서는 경사도가 증가할수록 증가하였으며 $10\% 경사일 때를 기준으로 $Ro(mm)=Ro_{10}{\times}0.797{\times}e^{-0.021s(\%)}$로 나타났다.천성 승모판 폐쇄 부전등을 초래하는 심각한 선천성 심질환이다. 그러나 진단 즉시 직접 좌관상동맥-대동맥 이식술로 수술적 교정을 해줌으로써 좋은 성적을 기대할 수 있음을 보여주었다.특히

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.14 no.6
• /
• pp.1-5
• /
• 2000

Recently alternating-current Plasma Display Panel(AC-PDP) is in the spotlight as a digital television and high definition television. The panel structure widely adapted in commercial AC-PDP is three electrodes surface discharge type. At present time, the luminous efficiency is around 1lm/W, it should be a key factor for the commercialization. For the high luminous efficiency, the development of panel structure is necessary. At a given panel structure, a driving method should be optimized to get a sufficient luminous efficiency. The display image of AC-PDP could be realized by the repeated light emission from the discharge. Because most of discharge power is consumed in the sustaining period, the optimization of sustaining waveform is very important for the high luminous efficiency. ADS (Address and Display period Separated) driving method is commonly used. The average driving frequency of ADS driving method is ranged by several tens kilo of [kHz], however the actual frequency of sustaining period is in range of 100[kHz] to 200[kHz]. Based on this study, when the phosphor emits the visible light, it has a decay time of few milliseconds due to the material transfer to the phosphor to emit the visible light. Consequently the luminous efficiency decreases in proportion to the driving frequency. It is found that the luminous efficiency could be significantly improved by the low frequency sustaining driving method.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.37 no.1
• /
• pp.33-40
• /
• 2024

The volume of shells, a prominent form of marine waste, is steadily increasing each year. However, a significant portion of these shells is either discarded or left near coastlines, posing environmental and social concerns. Utilizing shells as a substitute for traditional aggregates presents a potential solution, especially considering the diminishing availability of natural aggregates. This approach could effectively reduce transportation logistics costs, thereby promoting resource recycling. In this study, we explore the feasibility of employing wasted shell aggregates in 3D concrete printing technology for marine structures. Despite the advantages, it is observed that 3D printing concrete with wasted shells as aggregates results in lower strength compared to ordinary concrete, attributed to pores at the interface of shells and cement paste. Microstructure characterization becomes essential for evaluating mechanical properties. We conduct an analysis of the mechanical properties and microstructure of 3D printing concrete specimens incorporating wasted shells. Additionally, a mix design is proposed, taking into account flowability, extrudability, and buildability. To assess mechanical properties, compression and bonding strength specimens are fabricated using a 3D printer, and subsequent strength tests are conducted. Microstructure characteristics are analyzed through scanning electron microscope tests, providing high-resolution images. A histogram-based segmentation method is applied to segment pores, and porosity is compared based on the type of wasted shell. Pore characteristics are quantified using a probability function, establishing a correlation between the mechanical properties and microstructure characteristics of the specimens according to the type of wasted shell.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.34 no.6
• /
• pp.51-58
• /
• 2006

The spray characteristics of liquid jet injected into subsonic cross-flow were investigated experimentally. Spray trajectories were captured using CCD camera. Droplet sizes were measured using PDPA and Image Express. The nozzle diameter was 0.5 mm, and its length-to-diameter ratios (L/D) ran$4.11{\times}10^6$ged from 1.0 to 6.0. Experimental results indicate that the breakup point is delayed by increasing gas momentum ratio and the penetration length is decreased by increasing Weber number. At low injection angle(${\theta}$ < $90^{\circ}$), Weber number is dominant parameter for trajectories, but at high injection angle(${\theta}$ > $90^{\circ}$), L/D is dominant parameter for trajectories rather than Weber number.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2015.08a
• /
• pp.183-183
• /
• 2015

Indium tin oxide (ITO)는 넓은 밴드갭을 가지는 n-type의 축퇴 반도체로 태양전지, 스마트윈도우, 터치 센서, organic light emitting displays (OLEDs) 등에 널리 적용된다. 최근 touch screen panels (TSPs)의 높은 전기적 특성 및 고해상도 요구에 따라 고품질 ITO 박막개발의 수요도 증가하는 추세이다. 지금까지 ITO 박막의 물성 및 기계적 특성에 관한 많은 연구가 진행되어 왔지만 ITO 초박막 에서의 근본적인 물성 변화에 대한 연구는 미흡한 실정이므로, 이러한 연구는 필수적이라 할 수 있다. ITO 초박막은 광학적 특성은 우수하나, 낮은 결정성으로 인해 전기적 특성이 나쁘다는 단점을 가지며, 이러한 ITO 박막의 결정성은 초기 박막 성장과정에 많은 영향을 받는다. ITO 박막의 초기성장과정은 핵이 생성된 후(nucleation), 각각의 위치에서 성장하게 되고(growth), 합쳐지면서(coalescence) 연속적인 막을 형성 하는데(continuous), 이러한 초기 박막 성장 과정 중에 핵 생성 밀도를 증가시키고 박막이 연속적으로 되는 두께를 감소시킨다면, 더욱 더 고품질의 ITO 초박막을 얻을 수 있을 것이다. 따라서, 본 연구에서는 박막 초기 형성 과정 중 섬들이 합체되는 두께를 최소화시키기 위하여 EMF(electromagnetic field strength) 시스템을 이용하였다. EMF 시스템은 DC 캐소드에 전자석 코일을 장착하여 전자기장을 추가로 부가한 것으로, 이를 이용할 경우 스퍼터 원자가 중성상태로 기판에 도달하는 것이 아니라, 이온화되어 Vp-Vf의 차이로 가속되어 추가적인 에너지를 공급받음으로써 기판표면상에서 확산을 촉진시키므로 박막이 연속적으로 되는 임계 두께를 감소시킬 수 있는 것으로 기대된다. 실험은 실온에서 DC 마그네트론 스퍼터링법을 이용하였으며, 유리기판위에 4, 6, 8, 10, 12, 20 nm의 두께로 ITO 박막을 제작하였다. 스퍼터링 파워는 150 W (3.29 W/cm3), 작업 압력은 0.13 Pa, 기판과 타깃 사이의 거리는 70 mm였다. 각각의 두께에서 EMF 파워 0, 5, 10, 15, 20, 25, 30 W로 인가하여 박막을 제작한 후, EMF 파워에 따른 ITO 박막의 초기 성장 과정중 표면상태를 AFM (atomic force microscope) 이미지를 통하여 관찰하였다. 또한, 두께 약 8 nm에서와 20 nm일 때의 전기적 특성 및 광학적 특성을 관찰하였으며, 두 박막 모두 EMF 파워 15 W를 인가하였을 때 그 특성이 가장 향상되는 것을 확인하였다. 이러한 결과를 통하여 박막은 초기 성장이 중요하므로, 매우 얇은 두께에서 좋은 특성을 가진 박막을 제작하여야 박막의 두께를 증가시켰을 때도 좋은 특성의 막을 얻을 수 있음을 알 수 있었다. 또한, EMF 파워를 증가시킴에 따라 자장강도를 증가시키는 것과 같은 효과 즉, 플라즈마 임피던스가 감소하는 효과를 내어 증착 중 고 에너지 입자 (Ar0, O-)에 의한 박막손상이 감소한 것으로 판단된다. 따라서 적정 EMF 파워 15 W를 인가하였을때 가장 물성이 좋은 ITO 박막을 얻을 수 있었다. 즉, EMF 시스템을 이용하여 저온 공정에서 결함농도가 적은 고품질의 ITO 초박막을 제작할 수 있었다.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.43 no.6
• /
• pp.507-515
• /
• 2015

This paper proposes a filtering method for effective state vector estimation of highly maneuvering target. It is needed to hit the point called 'sweet spot' to increase the kill probability in missile interception. In paper, a filtering method estimates the length of a moving target tracked by a frequency modulated continuous wave (FMCW) radar. High resolution range profiles (HRRPs) is generated from the radar echo signal and then it's integrated into proposed filtering method. To simulate the radar measurement which is close to real, the study on the properties of scattering point of the missile-like target has been conducted with ISAR image for different angle. Also, it is hard to track the target efficiently with existing Kalman filters which has fixed measurement noise covariance matrix R. Therefore the proposed method continuously updates the covariance matrix R with sensor measurements and tracks the target. Numerical simulations on the proposed method shows reliable results under reasonable assumptions on the missile interception scenario.

• Journal of Korean Society for Geospatial Information Science
• /
• v.18 no.2
• /
• pp.3-12
• /
• 2010

In previous studies, the digital measurement systems and analysis algorithms were developed by using the related techniques, such as the aerial photograph detection and high resolution satellite image process. However, these studies were limited in 2-dimensional geo-processing. Therefore, it is necessary to apply the 3-dimensional spatial information and coordinate system for higher accuracy in recognizing and locating of geo-features. The objective of this study was to develop a stochastic algorithm for the optimal sensor placement using the 3-dimensional spatial analysis method. The 3-dimensional information of the LiDAR was applied in the sensor field algorithm based on 2- and/or 3-dimensional gridded points. This study was conducted with three case studies using the optimal sensor placement algorithms; the first case was based on 2-dimensional space without obstacles(2D-non obstacles), the second case was based on 2-dimensional space with obstacles(2D-obstacles), and lastly, the third case was based on 3-dimensional space with obstacles(3D-obstacles). Finally, this study suggested the methodology for the optimal sensor placement - especially, for ground-settled sensors - using the LiDAR data, and it showed the possibility of algorithm application in the information collection using sensors.