• Korean Journal of Remote Sensing
• /
• v.18 no.4
• /
• pp.221-231
• /
• 2002

In this study we applied a DEM generation algorithm developed in-house to satellite images at various resolution and discussed the results. We tested SPOT images at l0m resolution, EOC images at 6.6m and IKONOS images at 1m resolution. These images include the same urban area in Daejeon city. For camera model, we used Gupta & Hartley's(1997) DLT model for all three image sets. We carried out accuracy assessment using USGS DTED for SPOT and EOC and 23 check points for IKONOS. The assessment showed that SPOT DEM had about 38m RMS error, EOC DEM 12m RMS error and IKONOS DEM 6.5m RMS error. In terms of image resolution, SPOT and EOC DEM error corresponds to 2∼4 pixels where as IKONOS DEM error 6∼7pixels. IKONOS DEM contains more errors in pixels. However, in IKONOS DEM, individual buildings, apartments and major roads are identifiable. All three DEMs contained errors due to height discontinuity, occlusion and shadow. These experiments show that our algorithm can generate urban DEM from 1m resolution and that, however, we need to improve the algorithm to minimize effects of occlusion and building shadows on DEMs.

• Atmosphere
• /
• v.29 no.1
• /
• pp.61-74
• /
• 2019

To investigate the observational environment, sky line and skyview factor (SVF) are calculated using a digital elevation model (DEM; 10 m spatial resolution) and 3 dimensional (3D) sky image at radiation site, Gangneung-Wonju National University (GWNU). Solar radiation is calculated using GWNU solar radiation model with and without the sky line and the SVF retrieved from the 3D sky image and DEM. When compared with the maximum sky line elevation from Skyview, the result from 3D camera is higher by $3^{\circ}$ and that from DEM is lower by $7^{\circ}$. The SVF calculated from 3D camera, DEM and Skyview is 0.991, 0.998, and 0.993, respectively. When the solar path is analyzed using astronomical solar map with time, the sky line by 3D camera shield the direct solar radiation up to $14^{\circ}$ with solar altitude at winter solstice. The solar radiation is calculated with minutely, and monthly and annual accumulated using the GWNU model. During the summer and winter solstice, the GWNU radiation site is shielded from direct solar radiation by the west mountain 40 and 60 minutes before sunset, respectively. The monthly difference between plane and real surface is up to $29.18M\;m^{-2}$ with 3D camera in November, while that with DEM is $4.87M\;m^{-2}$ in January. The difference in the annual accumulated solar radiation is $208.50M\;m^{-2}$ (2.65%) and $47.96M\;m^{-2}$ (0.63%) with direct solar radiation and $30.93M\;m^{-2}$ (0.58%) and $3.84M\;m^{-2}$ (0.07%) with global solar radiation, respectively.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.39 no.3
• /
• pp.141-148
• /
• 2021

This paper presents a method to detect zebra-crossing using deep learning which combines SegNet and ResNet. For the blind, a safe crossing system is important to know exactly where the zebra-crossings are. Zebra-crossing detection by deep learning can be a good solution to this problem and robotic vision-based assistive technologies sprung up over the past few years, which focused on specific scene objects using monocular detectors. These traditional methods have achieved significant results with relatively long processing times, and enhanced the zebra-crossing perception to a large extent. However, running all detectors jointly incurs a long latency and becomes computationally prohibitive on wearable embedded systems. In this paper, we propose a model for fast and stable segmentation of zebra-crossing from captured images. The model is improved based on a combination of SegNet and ResNet and consists of three steps. First, the input image is subsampled to extract image features and the convolutional neural network of ResNet is modified to make it the new encoder. Second, through the SegNet original up-sampling network, the abstract features are restored to the original image size. Finally, the method classifies all pixels and calculates the accuracy of each pixel. The experimental results prove the efficiency of the modified semantic segmentation algorithm with a relatively high computing speed.

• Journal of Internet Computing and Services
• /
• v.24 no.4
• /
• pp.57-64
• /
• 2023

The ROK military faces a significant challenge in its vigilance mission due to demographic problems, particularly the current aging population and population cliff. This study demonstrates the crucial role of the 4th industrial revolution and its core artificial intelligence algorithm in maximizing work efficiency within the Command&Control room by mechanizing simple tasks. To achieve a fully developed military surveillance system, we have chosen multi-object tracking (MOT) technology as an essential artificial intelligence component, aligning with our goal of an intelligent and automated surveillance system. Additionally, we have prioritized data visualization and user interface to ensure system accessibility and efficiency. These complementary elements come together to form a cohesive software application. The CCTV video data for this study was collected from the CCTV cameras installed at the 1st and 2nd main gates of the 00 unit, with the cooperation by Command&Control room. Experimental results indicate that an intelligent and automated surveillance system enables the delivery of more information to the operators in the room. However, it is important to acknowledge the limitations of the developed software system in this study. By highlighting these limitations, we can present the future direction for the development of military surveillance systems.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2024.01a
• /
• pp.55-58
• /
• 2024

최근 각 기업의 AI 면접시스템 도입이 증가하고 있으며, AI 면접에 대한 실효성 논란 또한 많은 상황이다. 본 논문에서는 AI 면접 과정에서 지원자를 평가하는 방식을 시각, 음성, 자연어처리 3영역에서 구현함으로써, 면접 지원자를 다방면으로 분석 방법론의 적절성에 대해 평가하고자 한다. 첫째, 시각적 측면에서, 면접 지원자의 감정을 인식하기 위해, 합성곱 신경망(CNN) 기법을 활용해, 지원자 얼굴에서 6가지 감정을 인식했으며, 지원자가 카메라를 응시하고 있는지를 시계열로 도출하였다. 이를 통해 지원자가 면접에 임하는 태도와 특히 얼굴에서 드러나는 감정을 분석하는 데 주력했다. 둘째, 시각적 효과만으로 면접자의 태도를 파악하는 데 한계가 있기 때문에, 지원자 음성을 주파수로 환산해 특성을 추출하고, Bidirectional LSTM을 활용해 훈련해 지원자 음성에 따른 6가지 감정을 추출했다. 셋째, 지원자의 발언 내용과 관련해 맥락적 의미를 파악해 지원자의 상태를 파악하기 위해, 음성을 STT(Speech-to-Text) 기법을 이용하여 텍스트로 변환하고, 사용 단어의 빈도를 분석하여 지원자의 언어 습관을 파악했다. 이와 함께, 지원자의 발언 내용에 대한 감정 분석을 위해 KoBERT 모델을 적용했으며, 지원자의 성격, 태도, 직무에 대한 이해도를 파악하기 위해 객관적인 평가지표를 제작하여 적용했다. 논문의 분석 결과 AI 면접의 다면적 평가시스템의 적절성과 관련해, 시각화 부분에서는 상당 부분 정확도가 객관적으로 입증되었다고 판단된다. 음성에서 감정분석 분야는 면접자가 제한된 시간에 모든 유형의 감정을 드러내지 않고, 또 유사한 톤의 말이 진행되다 보니 특정 감정을 나타내는 주파수가 다소 집중되는 현상이 나타났다. 마지막으로 자연어처리 영역은 면접자의 발언에서 나오는 말투, 특정 단어의 빈도수를 넘어, 전체적인 맥락과 느낌을 이해할 수 있는 자연어처리 분석모델의 필요성이 더욱 커졌음을 판단했다.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.44 no.1
• /
• pp.1-9
• /
• 2024

Bridges, which are construction structures, have increased from 584 to 38,405 since the 1970s. However, as the number of bridges increases, the number of bridges with a service life of more than 30 years increases to 21,737 (71%) by 2030, resulting in fatal accidents due to basic human resource maintenance of facilities. Accordingly, the importance of bridge safety inspection and maintenance measures is increasing, and the need for decision-making support for supervisors who manage multiple bridges is also required. Currently, the safety inspection and maintenance method of bridges is to write down damage, condition, location, and specifications on the exterior survey map by hand or to record them by taking pictures with a camera. However, errors in notation of damage or defects or mistakes by supervisors are possible, typos, etc. may reduce the reliability of the overall safety inspection and diagnosis. To improve this, this study visualizes damage data recorded in the BIM model in an AR environment and proposes a maintenance plan for bridges with a small number of people through maintenance decision-making support for supervisors.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.35 no.2
• /
• pp.93-101
• /
• 1999

A simple experimental method was used in an attempt to realize the elevation of the fishing ability of purse seine in the sea area of Cheju Island, the changes of seine volume and tension in the purseline during pursing. Experiments carried out on the six types simplified reduced model seines which were made of knotless nettings. The nettings were woven in different leg length 4.3, 5.0, 5.5, 6.0, 6.6 and 7.7mm of polyester 28 tex two threads two-ply twine, and each of the seines were named I, II, III, IV, V and Ⅵ seine. Dimension of seine models were 450cm for corkline and 85cmfor seine depth, each seines rigged up 160g of float for a floatline and 50g (underwater weight) of lead for a leadline. These model purse seines were made of the scale of 1/200 of its full scale, a 120 ton in the near sea of Cheju Island. Designing and testing for the model purse seines were based on the Tauti's law. Experiments were measured in the observation channel of a flume tank at the static conditions set up shooting and pursing equipments. Motion of purse seine during purse line was recorded by the two sets video camera for VTR which were placed in top and front of the model seine. The reading coordinate of seine volume carried out by the video digitization system, disk data for the purseline tension. An analysis were performed on the changes seine volume and tension in the purseline during pursing. The results obtained were as follows: 1. The seine volume during pursing was largest for Ⅵ seine with smallest d/l followed by V, IV, III, II and I seines, and tension in the purseline was small. 2. Seine volume during pursing can be expressed by the following equation; CVt=l-EXP[｛2.79 (d/l）＋0.35｝t-33.37 (d/l) ＋ 0.57] Where CVt is volume ratio, d is twine diameter, l is leg length and t is pursing time (sec). 3. Tension in the purse line during pursing can be expressed by the following equation; T= 1- EXP ｛0.57t ＋ 13.36 (d/l)+2.97｝ Where T is tension (kg) in the purseline during pursing.

• Korean Journal of Applied Biomechanics
• /
• v.12 no.2
• /
• pp.17-32
• /
• 2002

The purpose of this study was to find the rational method to analyze golf swing with specific property of club shaft. Three subjects were filmed by two high speed digital cameras with 500 fps. The phase analyzed was downswing of each subject. The three-dimensional coordinates of the anatomical landmarks were obtained with motion analysis system Kwon3d 3.0 version and smoothed by lowpass digital filter with cutoff frequency 6Hz. From these data, kinematic and kinetic variables were calculated using Matlab(ver 5.0) The variables for this study were angular velocity and accelerations, which were calculated and following conclusions have been made : 1) Golf swing time of stiff club is faster than that of regular club. 2) In shoulder joint motion of swing with the stiff club, x-stiff showed mort rapid negative acceleration than that of regular club. 3) In regular club, the velocity of club head would be more effective velocity, which was increasing, than those of other clubs before impact. 4) In wrist joint motion of swing with stiff club, x-stiff club showed faster than regular club in the downswing and impact more rapid negative acceleration.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.37 no.2
• /
• pp.233-239
• /
• 2013

In this paper, a method of combining numerical analyses and experimental tests is used to evaluate fracture toughness of copper thin films of $15{\mu}m$ thickness. Far-field loadings of a global-local finite element model are inversely estimated by matching crack opening profiles in experiments with numerical results. The fracture toughness is then evaluated using the J-integral for cracks in thin films under far-field loadings. In experiments, Cu thin films attached to Aluminum sheets are loaded indirectly, and crack opening profiles are observed by microscope camera. Stress versus strain curves of Cu thin films are obtained through micro-tensile tests, and the grain size of Cu thin films is observed by TEM analysis. The results show that the fracture toughness of Cu thin films with $500nm{\sim}1{\mu}m$ sized grains is $6,962J/m^2$.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.23 no.3
• /
• pp.132-145
• /
• 2020

By coming of the 4th industrial revolution era, HD map have became a key infrastructure for determining precise location of autonomous driving in areas of futuristic cars, logistics and robots. Autonomous vehicles have became more dependent on HD map to determine the exact location of objects detected by various sensors such as LiDAR, GNSS, Radar, and stereo cameras as well as self-location decisions. By actualizing autonomous driving and C-ITS technologies, the demand for precise information on HD map have increased. And also the demand for the creation of new information based on the convergence of various changes and real-time information have increased. In this study, domestic and international HD map standards and related environments have analyzed. Based on this, usability has researched which comparison with standard HD map established by various institutions. Additionally, usability of standard HD map have studied for applying actual autonomous vehicles by reworking HD map. By the result of study, standard HD map have well established to use by various institutions. If further research about layer classification and definition by institutions will carried out based on this study, it has expected that and efficient establishment and renewal of HD map will take place.