• 한국도로학회논문집
• /
• 제19권6호
• /
• pp.147-153
• /
• 2017

PURPOSES : This research aims to estimate the occurrence of hydroplaning on roads based on the road alignment types and rainfall intensity in Seoul. METHODS : Three types of data were used for estimation of hydroplaning in this study. The Inner Circulation Road (12.5 km) to the Bukbu Expressway (7.4 km) in Seoul was selected as the test road and data was collected for road information using a probe-vehicle. Precipitation was observed from Automatic Weather System in Seoul. These data were interpolated by applying Inverse Distance Weighted Methodology for hydroplaning estimation. Finally, the water depth information of the roads was observed using an RCM411 device. RESULTS : This study demonstrated that the cross slope with small-angle-tilt or vertical section with large-angle-tilt are the primary factors causing hydroplaning on the roads. The flow velocity on steep slope is high; however, large drainage lengths result in hydroplaning on the roads. CONCLUSIONS : This result can contribute towards the reduction of car accidents on rainy days. Furthermore, information regarding hydroplaning can be delivered to drivers more rapidly and precisely in the future.

• 한국정보통신학회논문지
• /
• 제19권3호
• /
• pp.651-658
• /
• 2015

최근에 자율주행자동차에 대해 매우 활발한 연구와 개발이 진행되고 있다. 자율주행자동차를 구현하기 위해서는 매우 많은 기술들이 융복합적으로 해결되어야 한다. 이를 위해 차량에 장착된 블랙박스는 단순히 녹화기능 뿐만 아니라 신호등인식, 보행자검출, 정지선인식 등과 같이 자율주행차량을 구현하기 위한 핵심적인 기능을 제공할 수 있어 많은 연구 대상이 되고 있다. 따라서 자율주행차량을 구현하기 위한 한 가지 접근방법으로서 본 논문에서는 차량에 장착된 블랙박스 카메라를 이용하여 도로상에 위치한 거리를 효과적으로 예측할 수 있는 수식적인 모델을 제시한다. 제안한 모델은 도로의 기준선과 관찰선의 폭 또는 블랙박스 장착 높이 정보만을 이용함으로써 실제 도로상의 거리를 예측하는데 효과적으로 활용할 수 있음을 보인다. 다양한 실험을 통하여 본 논문에서 제안한 도로상의 거리 예측 모델이 타당함을 보인다.

• 한국통신학회논문지
• /
• 제38C권4호
• /
• pp.359-364
• /
• 2013

본 논문에서는 선박 접안 유도 시스템 등 정밀한 거리 추정을 요하는 안전항해운항, 선박트래픽관리 등의 분야에 알맞은 상관값 누적 기반 직접 수열 초광대역 (direct sequence ultra wideband: DS-UWB) 레이더 시스템을 제안한다. 제안하는 DS-UWB 레이더 시스템은 상관기 출력들을 누적하여 잡음을 평균냄으로써 잡음 성분의 영향을 줄이며, 기존의 DS-UWB 레이더 시스템들에 비해 더 빠른 시간에 신뢰성 있는 거리 추정을 수행한다. 모의실험 결과를 통해 제안하는 DS-UWB 레이더 시스템이 기존의 DS-UWB 레이더 시스템들에 비해 평균 상관 처리 시간이 더 짧을 뿐 아니라, 더 좋은 거리 추정 성능을 보임을 확인한다.

• Journal of Electrical Engineering and Technology
• /
• 제10권3호
• /
• pp.1144-1153
• /
• 2015

The resonant mechanism of reradiation interference (RRI) over 1.7MHz from power transmission lines cannot be obtained from IEEE standards, which are based on researches of field intensity. Hence, the resonance is ignored in National Standards of protecting distance between UHV power lines and radio stations in China, which would result in an excessive redundancy of protecting distance. Therefore, based on the generalized resonance theory, we proposed the idea of applying model-based parameter estimation (MBPE) to estimate the generalized resonance frequency of electrically large scattering objects. We also deduced equation expressions of the generalized resonance frequency and its quality factor Q in a lossy open electromagnetic system, i.e. an antenna-transmission line system in this paper. Taking the frequency band studied by IEEE and the frequency band over 1.7 MHz as object, we established three models of the RRI from transmission lines, namely the simplified line model, the tower line model considering cross arms and the line-surface mixed model. With the models, we calculated the scattering field of sampling points with equal intervals using method of moments, and then inferred expressions of Padé rational function. After calculating the zero-pole points of the Padé rational function, we eventually got the estimation of the RRI’s generalized resonant frequency. Our case studies indicate that the proposed estimation method is effective for predicting the generalized resonant frequency of RRI in medium frequency (MF, 0.3~3 MHz) band over 1.7 MHz, which expands the frequency band studied by IEEE.

• 대한임베디드공학회논문지
• /
• 제19권1호
• /
• pp.9-17
• /
• 2024

It is widely recognized that for 3D human pose estimation (HPE), dataset acquisition is expensive and the effectiveness of augmentation techniques of conventional visual recognition tasks is limited. We address these difficulties by presenting a simple but effective method that augments input images in terms of viewpoints when training a 3D human pose estimation (HPE) model. Our intuition is that meaningful variants of the input images for HPE could be obtained by viewing a human instance in the images from an arbitrary viewpoint different from that in the original images. The core idea is to synthesize new images that have self-occlusion and thus are difficult to predict at different viewpoints even with the same pose of the original example. We incorporate this idea into the training procedure of the 3D HPE model as an augmentation stage of the input samples. We show that a strategy for augmenting the synthesized example should be carefully designed in terms of the frequency of performing the augmentation and the selection of viewpoints for synthesizing the samples. To this end, we propose a new metric to measure the prediction difficulty of input images for 3D HPE in terms of the distance between corresponding keypoints on both sides of a human body. Extensive exploration of the space of augmentation probability choices and example selection according to the proposed distance metric leads to a performance gain of up to 6.2% on Human3.6M, the well-known pose estimation dataset.

• 한국음향학회지
• /
• 제28권4호
• /
• pp.383-391
• /
• 2009

음성 신호처리 환경에서 잡음이 섞인 신호를 개선할 목적으로 음성향상 기법이 많이 이용되고 있다. 잡음추정 알고리즘은 변화하는 환경에 빠르게 적응할 수 있어야 하며 음성신호의 영향을 줄이기 위해 음성신호가 존재하지 않는 구간에서만 잡음의 파워를 갱신한다. 이러한 방법은 음성구간검출이 선행되어야 한다. 그러나 잡음에 열화된 음성신호에 묵음구간이 존재하지 않을 경우, 위와 같이 음성검출을 통한 묵음구간에서의 잡음 추정 방법 및 SNR 추정 방법이 적용될 수 없다. 본 논문에서는 묵음구간이 존재하지 않는 연속음성신호에서 SNR을 추정하는 기법을 제안한다. 유성음의 안정구간에서는 단구간 내 피치의 변화가 매우 작아 피치주기에 따른 음성신호의 파형이 유사하게 나타난다. 따라서 잡음이 음성에 부가되었을 때 피치주기에 따른 인접파형의 유사도를 통해 SNR을 추정한다. 무성음에서는 잡음의 영향이 수신신호의 성도성분 추정에 영향을 미치기 때문에 잡음환경에서 추정된 성도성분과 수신신호 스펙트럼 간의 거리를 이용하여 SNR을 추정한다. 마지막으로, 음성신호의 에너지가 유성음에 대부분 분포하기 때문에, 부가성 잡음 환경에서 유성음의 에너지를 음성신호의 에너지로 근사화하여 SNR을 추정할 수 있다.

• 한국컴퓨터정보학회논문지
• /
• 제21권10호
• /
• pp.11-19
• /
• 2016

The size of a display is large, The form becoming various of that do not apply to previous methods of gaze tracking and if setup gaze-track-camera above display, can solve the problem of size or height of display. However, This method can not use of infrared illumination information of reflected cornea using previous methods. In this paper, Robust pupil detecting method for eye's occlusion, corner point of inner eye and center of pupil, and using the face pose information proposes a method for calculating the simply position of the gaze. In the proposed method, capture the frame for gaze tracking that according to position of person transform camera mode of wide or narrow angle. If detect the face exist in field of view(FOV) in wide mode of camera, transform narrow mode of camera calculating position of face. The frame captured in narrow mode of camera include gaze direction information of person in long distance. The method for calculating the gaze direction consist of face pose estimation and gaze direction calculating step. Face pose estimation is estimated by mapping between feature point of detected face and 3D model. To calculate gaze direction the first, perform ellipse detect using splitting from iris edge information of pupil and if occlusion of pupil, estimate position of pupil with deformable template. Then using center of pupil and corner point of inner eye, face pose information calculate gaze position at display. In the experiment, proposed gaze tracking algorithm in this paper solve the constraints that form of a display, to calculate effectively gaze direction of person in the long distance using single camera, demonstrate in experiments by distance.

• 한국전문물리치료학회지
• /
• 제17권3호
• /
• pp.1-10
• /
• 2010

The purpose of this study was to measure intra-rater and inter-rater reliability and range of motion for measurement of passive shoulder internal rotation range of motion and to compare anterior glide distance of humeral head in three methods. Fifty healthy subjects and fifty patients with shoulder musculoskeletal pain were recruited for this study. The subjects' passive shoulder internal rotation range of motion was measured by visual estimation, manual stabilization, and pressure biofeedback unit methods. In two trials, measurements were performed on each subject by two examiners. Intraclass correlation coefficient (ICC(3,1)) was used to determine the reliability of each measurement. The intra-rater reliability of the three methods was excellent (ICC=.77~.93) in both groups. The inter-rater reliability of the visual estimation method was poor (ICC=.20, .29), the manual scapular stabilization method was poor and fair (ICC=.09, .50), and the pressure biofeedback unit method was excellent (ICC .86, .75) in the experimental and control groups. In the experimental group, the difference of examined range of motion by each examiner was significant in the visual estimation method and manual scapular stabilization method, but there was an insignificant difference between the groups is the pressure biofeedback unit method. This result suggests that the intra-rater and inter-rater reliability of a pressure biofeedback unit was better than the other methods. The difference in distance of the anterior glide of humeral head was insignificant among all the methods. The pressure biofeedback unit method was the most reliable method, so it is proposed to be a new and reliable method to measure internal rotation range of motion.

• 한국음향학회지
• /
• 제16권2호
• /
• pp.71-79
• /
• 1997

본 논문에서는 FVQ-DHMM(fuzzy vector quantization-discrete hidden Markov model)에서 강인한 출력확률의 추정을 위해서 코드워드 종속 거리 정규화와 출력확률에 대한 instar 형태의 퍼지 평활화 방법을 제안한다. FVQ-DHMM은 DHMM의 변형된 모델로, 상태별 출력확률이 입력패턴에 대한 각 코드워드와의 가중치와 출력확률의 곱에 대한 합의 형태로 추정된다. FVQ-DHMM의 성능이 가중치 요소와 상태별 출력분포에 영향을 받으므로, 가중치 요소와 상태별 출력분포를 강인하게 추정하는 방법이 필요하게 된다. 실험결과, 제안된 코드워드 종속 거리 정규화(CDDN : codeword dependent distance normalization)를 적용한 방법이 기존의 FVQ-DHMM에 비해 24%의 오인식률 감소가 있었으며, 상태별 출력분포에 대해서 평활화를 적용한 경우 79%의 오식율을 감소 시킴을 알 수 있었다. 이러한 결과는 제안된 CDDN과 퍼지 평활화의 사용이 향상된 인식율을 얻는데 주요하며, 결과적으로 제안된 방법이 FVQ-HMM을 위한 강인한 출력확률의 추정을 위한 대안으로 유용함을 보여준다고 할 수 있다.

• 한국전산유체공학회지
• /
• 제22권1호
• /
• pp.37-42
• /
• 2017

Infrared signals are frequently used to detect objects exposed to wide variety of environmental conditions. Detection by infrared signature is accomplished by distinguishing objects by using the IR radiant contrast between objects and the background. There are several methods of estimating the IR radiant contrast. The inverse distance weighting method, which is one of the IR radiant contrast estimation method using the effect of distance from objects, is known to be an effective way to analyze radiant contrast for complex backgrounds. However this method has a disadvantage of requiring a long calculation time. In this study we propose a statistical method of estimating the IR radiant contrast by using randomly selected pixels of arbitrary number among background pixels to reduce calculation time. Some measured IR images in MWIR and LWIR regions are used to test the applicability of the method proposed and we found that the proposed method is very effective in determining the IR radiant contrast showing very rapid estimation with minar accuracy loss.