• 한국운동역학회지
• /
• 제18권1호
• /
• pp.53-62
• /
• 2008

본 연구는 플랫폼 다이빙 종목 205B를 3차원 영상분석을 통하여 운동학적 분석을 하였다. 총소요시간은 평균 $1.112{\pm}0.12s$로 나타났으며 여유 있는 연기를 수행하기 위해 많은 소요시간을 확보하여야 할 것으로 사료된다, 이지시 신체중심의 높이는 평균 88.8cm이며 수평거리는 144.8cm, 수평, 수직속도는 평균 1.08m/s, 1.3m/s로 나타났으며, 회전동작은 높은 위치에서 수행해야 하며, 수평거리는 짧아야 입수자세에 유리 할 것으로 판단된다. 하지분절의 속도변화는 이지 시 속도가 감소한 이후 회전이 이루어지는 시점까지 상승하며 이지 순간에는 회전보다는 점프동작에 더 큰 영향을 미치며, 신체중심을 축으로 회전동작으로 이어지고 속도가 상승하는 것으로 판단된다. 몸통각속도는 이지 시에 크게 나타났는데, 이는 상 방향으로 점프하여 회전에 필요한 힘을 얻기 위해 상체분절을 최대한 곧게 펴고 신체중심을 높게 유지하여 도약을 해야 효과적인 동작을 수행 할 수 있기 때문으로 사료된다.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송∙미디어공학회 2022년도 하계학술대회
• /
• pp.205-208
• /
• 2022

Multi-exposure high dynamic range (HDR) image reconstruction, the task of reconstructing an HDR image from multiple low dynamic range (LDR) images in a dynamic scene, often produces ghosting artifacts caused by camera motion and moving objects and also cannot deal with washed-out regions due to over or under-exposures. While there has been many deep-learning-based methods with motion estimation to alleviate these problems, they still have limitations for severely moving scenes. They also require large parameter counts, especially in the case of state-of-the-art methods that employ attention modules. To address these issues, we propose a frequency domain approach based on the idea that the transform domain coefficients inherently involve the global information from whole image pixels to cope with large motions. Specifically we adopt Residual Fast Fourier Transform (RFFT) blocks, which allows for global interactions of pixels. Moreover, we also employ Depthwise Overparametrized convolution (DO-conv) blocks, a convolution in which each input channel is convolved with its own 2D kernel, for faster convergence and performance gains. We call this LFFNet (Lightweight Frequency Fusion Network), and experiments on the benchmarks show reduced ghosting artifacts and improved performance up to 0.6dB tonemapped PSNR compared to recent state-of-the-art methods. Our architecture also requires fewer parameters and converges faster in training.

• Earthquakes and Structures
• /
• 제5권2호
• /
• pp.161-205
• /
• 2013

We study seismically induced, anti-plane strain wave motion in a non-homogeneous geological region containing tunnels. Two different scenarios are considered: (a) The first models two tunnels in a finite geological region embedded within a laterally inhomogeneous, layered geological profile containing a seismic source. For this case, labelled as the first boundary-value problem (BVP 1), an efficient hybrid technique comprising the finite difference method (FDM) and the boundary element method (BEM) is developed and applied. Since the later method is based on the frequency-dependent fundamental solution of elastodynamics, the hybrid technique is defined in the frequency domain. Then, an inverse fast Fourier transformation (FFT) is used to recover time histories; (b) The second models a finite region with two tunnels, is embedded in a homogeneous half-plane, and is subjected to incident, time-harmonic SH-waves. This case, labelled as the second boundary-value problem (BVP 2), considers complex soil properties such as anisotropy, continuous inhomogeneity and poroelasticity. The computational approach is now the BEM alone, since solution of the surrounding half plane by the FDM is unnecessary. In sum, the hybrid FDM-BEM technique is able to quantify dependence of the signals that develop at the free surface to the following key parameters: seismic source properties and heterogeneous structure of the wave path (the FDM component) and near-surface geological deposits containing discontinuities in the form of tunnels (the BEM component). Finally, the hybrid technique is used for evaluating the seismic wave field that develops within a key geological cross-section of the Metro construction project in Thessaloniki, Greece, which includes the important Roman-era historical monument of Rotunda dating from the 3rd century A.D.