http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
- 中文 을 입력하시려면 zhongwen을 입력하시고 space를누르시면됩니다.
- 北京 을 입력하시려면 beijing을 입력하시고 space를 누르시면 됩니다.
RISS 인기검색어
- 검색키워드
- 저자 : Hogeon (검색결과 32 건)
- 무료
- 기관 내 무료
- 유료
-
Hogeon Seo,Dong-Gi Song,Kyung-Young Jhang 한국비파괴검사학회 2016 한국비파괴검사학회지 Vol.36 No.2
Measurement of elastic constants is crucial for engineering aspects of predicting the behavior of materials under load as well as structural health monitoring of material degradation. Ultrasonic velocity measurement for material properties has been broadly used as a nondestructive evaluation method for material characterization. In particular, pulse-echo method has been extensively utilized as it is not only simple but also effective when only one side of the inspected objects is accessible. However, the conventional technique in this approach measures longitudinal and shear waves individually to obtain their velocities. This produces a set of two data for each measurement. This paper proposes a simultaneous sensing system of longitudinal waves and shear waves for elastic constant measurement. The proposed system senses both these waves simultaneously as a single overlapped signal, which is then analyzed to calculate both the ultrasonic velocities for obtaining elastic constants. Therefore, this system requires just half the number of data to obtain elastic constants compared to the conventional individual measurement. The results of the proposed simultaneous measurement had smaller standard deviations than those in the individual measurement. These results validate that the proposed approach improves the efficiency and reliability of ultrasonic elastic constant measurement by reducing the complexity of the measurement system, its operating procedures, and the number of data.
-
Hogeon Seo,Sungmoon Joo 한국비파괴검사학회 2021 한국비파괴검사학회지 Vol.41 No.1
Laser scanning is a noncontact and nondestructive technique that captures the three-dimensional (3D) shape of objects as point clouds. Deep neural networks have been widely used to classify the 3D shapes of point clouds. In applying deep learning on point clouds, point cloud preprocessing is the first step. This study was conducted to analyze 3D shape classification characteristics using a deep neural network, PointNet, with a point cloud dataset, ModelNet40, for four preprocessing cases: random, scaling, zero-mean, and normalization. For each preprocessing case, the minimum and maximum coordinates of the point clouds and 3D shape classification performance are investigated. The results show that normalization preprocessing exhibits the most significant improvement in classification performance, and the zero-mean method is particularly effective. The findings indicate that proper preprocessing, such as normalization, should be performed before deep learning when the mean coordinates and scale of the point clouds differ significantly.
-
Hogeon Seo,Sungmoon Joo 제어로봇시스템학회 2020 제어로봇시스템학회 국제학술대회 논문집 Vol.2020 No.10
Three-dimensional (3D) laser scanning is widely used to acquire the structural information of a target as a point cloud and reconstruct its shape. Recently, deep learning has shown good performance for 3D point cloud shape classification. The preprocessing of the point cloud is a primary step of deep learning. This study presents the performance of 3D shape classification via PointNet with a point cloud dataset, ModelNet40, with respect to three preprocessing cases: Random, zero mean, and normalization. The minimum and maximum values of the point cloud are compared according to the preprocessing method. In training, the number of points as an input was 1024. In addition, the influence of two augmentation methods (i.e., resampling and zero filling) was investigated. For this, the number of points was increased to 2048. Of the 2048 points, 1024 points were used the same as in the previous experiment, while the remaining 1024 points were added by resampling or zero filling. The results show that the zero mean method is effective for deep learning and normalization is better, whereas increasing the input size with the resampling or zero filling rather degrades the performance and increases unnecessary training costs.
-
Numerical analysis on the aerodynamics of HAWTs using nonlinear vortex strength correction
Hogeon Kim,Seoungmin Lee,이수갑 한국물리학회 2010 Current Applied Physics Vol.10 No.2
Nonlinear vortex strength correction method (NVCM) based on potential flow, is developed for improvement of vortex lattice method which has difficulties to predict the separated flow conditions and the viscous effect. In this method, the bound vortex strength is determined by matching the lift force from VLM with the lift force from aerodynamic coefficients table as the same value of circulation is added to or subtracted from all chord wise vortices. For considering the nonlinearities due to the neighboring sections of the blade, sophisticated Newton–Rapson algorithm is applied. The validation of this method was done by comparing the simulations with the measurements on the NREL Phase-VI horizontal axis wind turbine (HWAT) in the NASA Ames wind tunnel under uniform and yawed flow conditions. This method gives good agreements with experiments in most cases.
-
-
Hogeon Seo,Myunghwan Kim,Sungho Choi,Chung Seok Kim,Kyung-Young Jhang 한국비파괴검사학회 2012 한국비파괴검사학회지 Vol.32 No.4
Using a single-line pulsed laser beam is well known as a useful noncontact method to generate a directional surface acoustic wave. In this method, different laser beam energy profiles produce different waveforms and frequency characteristics. In this paper, we considered two typical kinds of laser beam energy profiles, Gaussian and square-like, to find out a difference in the frequency characteristics. To achieve this, mathematical models were proposed first for Gaussian laser beam profile and square-like respectively, both of which depended on the laser beam width. To verify the theoretical models, experimental setups with a cylindrical lens and a line-slit mask were respectively designed to produce a line laser beam with Gaussian spatial energy profile and square-like. The frequency responses of the theoretical models showed good agreement with experimental results in terms of the existence of harmonic frequency components and the shift of the first peak frequencies to low.
-
-
-
Seo, Hogeon,Song, Dong-Gi,Jhang, Kyung-Young The Korean Society for Nondestructive Testing 2016 한국비파괴검사학회지 Vol.36 No.2
Measurement of elastic constants is crucial for engineering aspects of predicting the behavior of materials under load as well as structural health monitoring of material degradation. Ultrasonic velocity measurement for material properties has been broadly used as a nondestructive evaluation method for material characterization. In particular, pulse-echo method has been extensively utilized as it is not only simple but also effective when only one side of the inspected objects is accessible. However, the conventional technique in this approach measures longitudinal and shear waves individually to obtain their velocities. This produces a set of two data for each measurement. This paper proposes a simultaneous sensing system of longitudinal waves and shear waves for elastic constant measurement. The proposed system senses both these waves simultaneously as a single overlapped signal, which is then analyzed to calculate both the ultrasonic velocities for obtaining elastic constants. Therefore, this system requires just half the number of data to obtain elastic constants compared to the conventional individual measurement. The results of the proposed simultaneous measurement had smaller standard deviations than those in the individual measurement. These results validate that the proposed approach improves the efficiency and reliability of ultrasonic elastic constant measurement by reducing the complexity of the measurement system, its operating procedures, and the number of data.
-
포인트 클라우드 및 투영 이미지를 이용한 다중 모달 형상 분류
서호건(Hogeon Seo) 한국비파괴검사학회 2022 한국비파괴검사학회지 Vol.42 No.1
포인트 클라우드의 형상을 분류하기 위해 심층신경망을 활용할 때, 점들의 좌푯값만을 활용하거나 연산 부담이 큰 3차원 렌더링을 통해 생성한 이미지를 활용하여 형상 분류를 수행한다. 본 연구에서는 좌푯 값과 해당 좌표를 활용해 생성한 투영 이미지를 함께 다중 모달로 심층신경망의 입력으로 활용하는 포인트 클라우드의 형상 분류 기법을 제안한다. 성능 향상 여부를 확인하고자, 좌푯값 기반으로 형상을 분류하는 PointNet과 이미지 기반 분류 모델인 ResNet-18을 조합하여 다중 모달 모델을 구성하고 ModelNet40 데이터셋에 대해서 투영 이미지의 여부 및 방향(등각면, 정면, 측면, 상면)에 따른 성능 평가를 수행하였다. 그 결과 측면 투영 이미지가 함께 고려될 때 가장 성능이 좋았으며, 정면 투영 이미지의 경우가 두 번째로 우수한 성능을 보였다. 이는 포인트 클라우드의 형상 분류에 있어서 좌푯값과 더불어 투영 이미지를 함께 입력으로 활용하는 것이 형상 분류에 효과적임을 뒷받침한다. A deep neural network is used to classify the shape of the point cloud using only the coordinate values of the points or the rendered image. In this study, we proposed a point cloud shape classification technique using the coordinate values and the projection image generated using the coordinates as an input to a multi-modal deep neural network. To verify the performance improvement, the multi-modal model built using PointNet, which classifies the shapes based on coordinate values, and ResNet-18, an image-based classification model, was evaluated for the shape classification performance on a ModelNet40 dataset. The result showed that the performance was the best when the side projection image was additionally considered and the second best when the front projection image was considered. This supports the idea that the shape classification performance of the point cloud can be improved using the coordinate values and its projection image as the input to the deep neural network in a multi-modal manner.
연관 검색어 추천
이 검색어로 많이 본 자료
활용도 높은 자료
