• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.10
• /
• pp.419-426
• /
• 2022

In this paper, we propose a method to improve the accuracy of 3D human pose estimation model in various move motions. Existing human pose estimation models have some problems of jitter, inversion, swap, miss that cause miss coordinates when estimating human poses. These problems cause low accuracy of pose estimation models to detect exact coordinates of human poses. We propose a method that consists of detection and correction methods to handle with these problems. Deep learning-based outlier detection method detects outlier of human pose coordinates in move motion effectively and rule-based correction method corrects the outlier according to a simple rule. We have shown that the proposed method is effective in various motions with the experiments using 2D golf swing motion data and have shown the possibility of expansion from 2D to 3D coordinates.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.9 no.2
• /
• pp.159-172
• /
• 2006

In this study, the series of Landsat TM/ETM+ images was acquired to extract land surface temperature for wide-area and executed geometric correction and radiometric correction. And the land surface temperature was extracted using NASA Model, and achieved the first correction by performing land coverage category for study area and applied characteristic emission rate. Land surface temperature which was acquired by the first correction was analyzed in correlation with Meteorological Administration's temperature data by regression analysis, and established correction formula. And I wished to improve accuracy of land surface temperature extraction using satellite image by second correcting deviations between two data using establishing correction formula. As a result, land surface temperature acquired by 1st and 2st correction could be corrected in mean deviation of about ${\pm}3.0^{\circ}C$ with Meteorological Administration data. Also, I could acquire land surface temperature about study area by higher accuracy by applying to other Landsat images for re-verification of study results.

• Geophysics and Geophysical Exploration
• /
• v.13 no.2
• /
• pp.169-174
• /
• 2010

We developed a precise terrain correction program using triangular element method (TEM) for microgravity data processing. TEM calculates gravity attraction of arbitrary polyhedra whose surface is patched by triangles. We showed that TEM can calculate more precise terrain effect than conventional rectangular prism method. We tested the accuracy of TEM on the cone model which has analytic solution. Also, we tested the accuracy of TEM on the slope model, this results showed that there are big differences calculated by TEM and rectangular prsim method (RPM) on slope model. The developed terrain correction program was applied on the gravity data on the southern area near sea shore of Korean peninsula, calculated terrain effect very precisely.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.33 no.4
• /
• pp.47-56
• /
• 2005

In GPS/INS/barometer navigation system for UAV, vertical channel damping loop was introduced to suppress divergence of the vertical axis error of INS, which could be reduced to the level of accuracy of pressure altitude measured by a pitot-static tube. Because static pressure measured by the pitot-static tube depends on the speed and attitude of the vehicle, static pressure error models, based on aerodynamic data from wind tunnel test, CFD analysis, and flight test, were applied to reduce the error of pressure altitude. Through flight tests and sensitivity analyses, the error model using the ratio of differential pressure and static pressure turned out to be superior to the model using only differential pressure, especially in case of high altitude flight. Both models were proposed to compensate the effect of vehicle speed change and used differential and static pressure which could be obtained directly from the output of pressure transducer.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.23 no.2
• /
• pp.103-111
• /
• 2023

This paper proposes a new data correction technique that transforms anomalies in time series data into normal values. With the recent development of IT technology, a vast amount of time-series data is being collected through sensors. However, due to sensor failures and abnormal environments, most of time-series data contain a lot of anomalies. If we build a predictive model using original data containing anomalies as it is, we cannot expect highly reliable predictive performance. Therefore, we utilizes the LSTM-GAN model to detect anomalies in the original time series data, and combines DTW (Dynamic Time Warping) and GAN techniques to replace the anomaly data with normal data in partitioned window units. The basic idea is to construct a GAN model serially by applying the statistical information of the window with normal distribution data adjacent to the window containing the detected anomalies to the DTW so as to generate normal time-series data. Through experiments using open NAB data, we empirically prove that our proposed method outperforms the conventional two correction methods.

• Journal of Korea Water Resources Association
• /
• v.37 no.9
• /
• pp.719-727
• /
• 2004

Water quality models can be applied to manage the regional water quality problems and to estimate the target and allowable pollution load in watershed effectively. The optimization of state variables in the given water quality model Is necessary to build up more effective model. The least-squares method is applied to fit field observations in QUAL2E developed by U.S. EPA, which is most widely used one in the world to simulate the stream water quality, and the optimization model with constraints is constructed to estimate the parameters. The objective function of the optimization model is solved by Solver in Microsoft Excel and Monte Carlo simulation is conducted to know the influence of parameter in conventional pollutants. It is found that this technique is easily implemented and rapidly convergent computational procedure to calibrate the parameters after appling this approach in Anyang stream located in Kyonggi province mainly.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2013.10a
• /
• pp.432-434
• /
• 2013

본 논문의 실내위치측위시스템은 무선랜 환경에서 AP의 RSSI 신호를 수신하여 Friis 공식을 통해 거리로 산출되고, 산출된 거리는 삼각측량법에 의해 (x, y)좌표로 변환되어 현재의 위치를 나타낸다. 여기서 RSSI 신호는 신호잡음을 포함하고 신호잡음으로 인하여 실제 위치 측위 시 오차가 발생한다. 이러한, 오차를 보정하고 실내위치측위 정확도를 향상하기 위해 비선형시스템에서 사용하는 확장칼만필터를 적용하여 실험하였다. 본 실내위치측위시스템의 시스템모델은 선형이고 측정모델은 비선형이므로 효율적인 보정알고리즘인 확장칼만필터를 선택하고 실험은 MATLAB로 수행하였다. 실험결과 실내위치측위시스템의 정확도가 향상되었다.

• Korean Journal of Optics and Photonics
• /
• v.16 no.2
• /
• pp.128-132
• /
• 2005

Most CGH programs use a model equation based on the diffraction angle. Therefore, if the diffraction angle is large enough, the image on a flat screen is distorted. To correct the distortion, we created the model equation from diffraction theory and verified it through experiment. We also suggest a design method that compensates for the distortion without changing the CGH program.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.20 no.2
• /
• pp.155-164
• /
• 2002

In this study, in order to discuss the geometric correction methods of high-resolution IKONOS satellite image, the existing polynomial model and RFM which is able to rectify satellite image without auxiliary data are applied to IKONOS satellite image data. Then the accuracy of ground point versus number of GCPs and each order of RFM are assessed. A numerical instability is removed by application of Tikhonov regularization method. As the results of this study, the root mean square errors of RFM is decreased more than 2 pixels in comparison with the two dimensional polynomial model.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2006.06a
• /
• pp.5-8
• /
• 2006

본 논문에서는 3차원 좌표를 얻을 수 있는 카메라 Calibration 알고리듬을 확립하고, View Frustum(V.F.) 모델을 이용하여 도로의 영상을 모델화하였다. 그리고 주행하는 차선 내에 존재하는 선행차량의 위치측정 및 차량까지의 거리를 정확히 인식하기 위해 피칭오차를 보정하며 실시간으로 계측하는 알고리듬을 제안하였다. 기존의 많은 추돌 경보시스템(CWS)들은 도로가 평면이라 가정하여 도로와 차량사이의 기하적인 변화에 따른 오차 특성을 고려하지 않았다. 이를 보완하고자 본 논문에서는 카메라 Calibration 알고리듬을 적용하여 실세계 좌표계와 영상좌표계 사이의 기하해석으로 사영행렬을 추출하였고, V.F. 모델을 이용하여 소실점의 기하적인 해석을 통하여 차량의 피칭변화에 따른 오차특성을 실시간으로 보정하였다. 실험결과 거리의 오차를 2%이하로 줄일 수 있어 피칭변화에 강인함을 확인할 수 있었다.