• Journal of Advanced Navigation Technology
• /
• v.16 no.6
• /
• pp.910-919
• /
• 2012

In this paper, we described development of on-board image processing system and its process and verified its performance through flight experiment. The image processing board has single ARM(Advanced Risk Machine) processor. We performed Embedded Linux Porting. Algorithm to be applied for object tracking is color-based image processing algorithm, it can be designed to track the object that has specific color on ground in real-time. To verify performance of the on-board image processing system, we performed flight test using the PNUAV, UAV developed by LAB. Also, we performed optimization of the image processing algorithm and kernel to improve real-time performance. Finally we confirmed that proposed system can track the blue-color object within four pixels error range consistently in the experiment.

• Journal of Energy Engineering
• /
• v.25 no.1
• /
• pp.86-91
• /
• 2016

Convenient design guideline for Waveguide-below-cutoff (WBC) array is proposed to obtain the minimum waveguide length for electromagnetic pulse (EMP) shielding. The analysis includes circular, rectangular, and hexagonal WBC, determine the total length of the waveguide. When the unit side of rectangular WBC and the diagonal line of hexagonal WBC are given as 30 mm, the length of hexagonal WBC is 5 mm shorter than rectangular case with shielding effectiveness (SE) 80 dB. The length difference is deepened with SE of 100 dB, which shows approximately 30 mm shorter length for hexagonal case than others. In addition, hexagonal WBC requires much shorter length than circular WBC. In conclusion, hexagonal case is the most effective with respect to flow velocity and pressure loss for equivalent SE.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.19 no.1
• /
• pp.61-68
• /
• 2015

Noise reduction is necessary to compensate for the degradation of recognition performance by various types of noises. Among many noise reduction techniques using microphone array, generalized sidelobe canceller (GSC) has been widely applied to reduce nonstationary noises. The performance of GSC is directly affected by its adaptation mode controller (AMC). That is, accurate target speech detection is essential to guarantee the sufficient noise reduction in pure noise intervals and the less distortion in target speech intervals. Thus, this paper proposes an improved AMC design technique in which the power ratio of the output of fixed beamforming to that of blocking matrix is calculated frequency bandwise and probabilistically modeled by mixture Gaussians for each class. Experimental results show that the proposed algorithm outperforms conventional AMCs in receiver operating curves (ROC) and output SNRs.

• The Korean Journal of Applied Statistics
• /
• v.34 no.2
• /
• pp.167-175
• /
• 2021

Semi-supervised learning uses both labeled data and unlabeled data. Recently consistency regularization is very popular in semi-supervised learning. Unsupervised data augmentation (UDA) that uses unlabeled data augmentation is also based on the consistency regularization. The Kullback-Leibler divergence is used for the loss of unlabeled data and cross-entropy for the loss of labeled data through UDA learning. UDA uses techniques such as training signal annealing (TSA) and confidence-based masking to promote performance. In this study, we propose to use Jensen-Shannon divergence instead of Kullback-Leibler divergence, reverse-TSA and not to use confidence-based masking for performance improvement. Through experiment, we show that the proposed technique yields better performance than those of UDA.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.365-365
• /
• 2023

우리나라는 지리, 지형 및 기상특성으로 인해 집중호우와 태풍 등으로 자연재해 피해를 입고 있으며, 재해예방사업의 일환으로 1998년부터 자연재해위험개선지구 정비사업을 통해 자연재해로부터 침수, 유실, 붕괴위험 등 피해예방을 위해 노력하고 있다. 서울특별시는 도시화로 인한 우수침투 가능지역 감소와 유역경사 부족에 따른 충분한 관거 경사확보의 어려움으로 저지대에 노면수가 집중되면서 대규모 침수피해(2010. 09, 2011. 07)가 발생되었으며, 신월·신정지구 및 화곡2지구가 자연재해위험개선지구로 선정(2011. 04), 침수방지를 위한 대책으로 국내 최초의 대규모 터널형 빗물저류배수시설(신월빗물저류배수시설)을 준공(2020. 05)하였다. 본 연구에서는 신월빗물저류배수시설 설치에 따른 상습침수구역의 사업 전·후 침수피해 저감효과 검증을 위해 도시유출모형을 구축하여 서울시 방재성능목표 확보여부를 판단하고자 하였다. 기존 설계 당시 서울시 방재성능목표(30년 빈도, 95mm/hr)는 확보되었으나 이를 초과하는 강우(2010. 09, 98.5mm/hr)가 발생함에 따라 수행된 추가 분석(100mm/hr) 결과, 유역의 침수 완전해소를 위해서는 빗물저류배수시설과 더불어 국부적인 간선관거 개량의 필요성이 제시되었다. 이에 따라 본 연구에서는 도시유출모형 구축 시 최근 개선된 하수관로, 지형특성, 홍수량산정표준침(2019, 환경부)등에 따른 수리·수문특성을 반영하고, 신일펌프장(간이)을 추가적으로 고려하여 현재 기준에서의 침수피해 저감 효과를 면밀이 분석하였으며, 그 결과 서울시 방재성능목표 뿐만 아니라 100mm/hr에도 침수가 완전 해소되는 것으로 나타났다. 신월빗물저류배수터널 설치에 따른 침수피해 저감효과 평가 및 분석결과를 통해 자연재해위험개선지구의 해제가 가능할 것으로 판단되며, 지역안전·환경효과, 주민만족도 등 다양한 항목에 대하여 상습침수구역 내 주민설문조사를 수행한 결과 신월빗물저류배수터널에 대한 만족도가 충분히 제고된 만큼 향후 시설의 기능적, 운영적 측면에서의 유지관리가 잘 이루어진다면 침수로부터 안전한 도시를 구현하는데 큰 역할을 할 것으로 기대된다.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.3
• /
• pp.212-218
• /
• 2013

This paper presents an SVD(Singular Value Decomposition)-Pseudo Spectrum method for SAR (Synthetic Aperture Radar) imaging. The purpose of this work is to improve resolution and target separability of SAR images. This paper proposes SVD-Pseudo Spectrum method whose advantages are noise robustness, reduction of sidelobes and high resolution of spectral estimation. SVD-Pseudo Spectrum method uses Hankel Matrix of signal components and SVD (Singular Value Decomposition) method. In this paper, it is demonstrated that the SVD-Pseudo Spectrum method shows better performance than the matched filtering method and the conventional super-resolution based multiple signal classification (MUSIC) method in SAR image processing. The targets to be separated are modeled, and this modeled data is used to demonstrate the performance of algorithms.

• The Journal of the Acoustical Society of Korea
• /
• v.20 no.8
• /
• pp.44-57
• /
• 2001

This paper analyses the performance of FFTSA (Fast Fourier Transform Synthetic Aperture) in the effects of temporal coherence and oscillatory towed course, which is one of the techniques for passive synthetic aperture SONAR process using linearly distributed towed array. Also this paper proposes the FFTSA technique using towed array having conformal shape to alleviate the performance degradation for estimating the incident angle under inconsistent under water environments. And this paper analyses the performance of the proposed FFTSA technique making use of conformal structure throughout exhaustive computer simulations.

• Journal of the Society of Naval Architects of Korea
• /
• v.32 no.3
• /
• pp.19-28
• /
• 1995

This paper presents hull form design details and performance characteristics of the SWATH ship 'SEON JIN' built as an underwater test support ship. The hull form is developed systematically varing sets of 11 form parameters which define a unique hull under waterline. Using this geometry variation scheme we have generated a number of alternatives and selected the final hull which fulfills the design objectives and matches the design constraints. After selecting the final hull form we have investigated performances thoroughly through model tests, and confirmed through sea trials that the performance goals be achieved.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2013.10a
• /
• pp.750-753
• /
• 2013

In respect of consuming energy, the optimization is the main objective in the solar energy harvesting sensor system (while battery-based sensor system aims at the minimization), due to the periodicity of solar energy. Aimed at the optimization of the network topology, we suggest 3-level transmission power control algorithm of which level is determined by the amount of residual energy on the rechargeable battery. Additionally, we experiment the effects of our scheme on network-wide performance such as the latency and the duty-cycle, and verify that our scheme shows the best performance in most of the metrics, compared to the schemes with fixed transmission power.

• The KIPS Transactions:PartB
• /
• v.15B no.1
• /
• pp.53-60
• /
• 2008

This paper proposes a new multiobjective optimization method for discriminative training of hidden Markov models (HMMs) used as the recognizer for automatic lipreading. While the conventional Baum-Welch algorithm for training HMMs aims at maximizing the probability of the data of a class from the corresponding HMM, we define a new training criterion composed of two minimization objectives and develop a global optimization method of the criterion based on simulated annealing. The result of a speaker-dependent recognition experiment shows that the proposed method improves performance by the relative error reduction rate of about 8% in comparison to the Baum-Welch algorithm.