• 한국멀티미디어학회논문지
• /
• 제20권11호
• /
• pp.1729-1740
• /
• 2017

This paper describes a vision-based method that effectively recognize a traffic light. The method consists of two steps of traffic light detection and discrimination. Many related studies have used color information to detect traffic light, but color information is not robust to the varying illumination environment. This paper proposes a new method of traffic light detection based on intensity and saturation. When a traffic light is turned on, the light region usually shows values with high saturation and high intensity. However, when the light region is oversaturated, the region shows values of low saturation and high intensity. So this study proposes a method to be able to detect a traffic light under these conditions. After detecting a traffic light, it estimates the size of the body region including the traffic light and extracts the body region. The body region is compared with five models which represent specific traffic signals, then the region is discriminated as one of the five models or rejected as none of them. Experimental results show the performance of traffic light detection reporting the precision of 97.2%, the recall of 95.8%, and correct recognition rate of 94.3%. These results shows that the proposed method is effective.

• Wind and Structures
• /
• 제27권4호
• /
• pp.255-267
• /
• 2018

This paper investigated the wind-snow flow around the bogie region of a high-speed train under crosswinds using a coupled numerical method of the unsteady Realizable $k-{\varepsilon}$ turbulence model and discrete phase model (DPM). The flow features around the bogie region were discussed and the influence of bogie fairing height on the snow accumulation on the bogie was also analyzed. Here the high-speed train was running at a speed of 200 km/h in a natural environment with the crosswind speed of 15 m/s. The mesh resolution and methodology for CFD analysis were validated against wind tunnel experiments. The results show that large negative pressure occurs locally on the bottom of wheels, electric motors, gear covers, while the positive pressure occurs locally on those windward surfaces. The airflow travels through the complex bogie and flows towards the rear bogie plate, causing a backflow in the upper space of the bogie region. The snow particles mainly accumulate on the wheels, electric motors, windward sides of gear covers, side fairings and back plate of the bogie. Longer side fairings increase the snow accumulation on the bogie, especially on the back plate, side fairings and brake clamps. However, the fairing height shows little impact on snow accumulation on the upper region of the bogie. Compared to short side fairings, a full length side fairing model contributes to more than two times of snow accumulation on the brake clamps, and more than 20% on the whole bogie.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제55권2호
• /
• pp.52-61
• /
• 2006

For the detection of moving objects, background subtraction methods are widely used. An adaptive Gaussian mixture model combined with probabilistic learning is one of the most popular methods for the real-time update of the complex and dynamic background. However, probabilistic learning approach does not work well in high traffic regions. In this paper, we Propose a reliable learning method of complex and dynamic backgrounds in high traffic regions.

• 한국산학기술학회논문지
• /
• 제16권2호
• /
• pp.1524-1532
• /
• 2015

본 연구에서는 우리나라 교통안전수준의 제고를 위해 충청권을 대상으로 교통사고분석자료를 이용한 요인분석 및 군집분석과 설문조사를 실시하여 충청권 교통사고 특성을 분석하였으며, 분석결과를 토대로 충청권 실정에 맞는 교통안전기술을 개발하였다. 충청권 운송업자들을 대상으로 실시된 설문조사결과, 충청권의 주요사고 위협요인으로 '보행자 및 고령자 통행'과 '시인성 부족문제'가 가장 큰 것으로 나타났고, 가장 많이 필요로 하는 시설물로서는 '야간 운전시 조명시설'인 것으로 나타났으며, 시설물 설치로 인한 기대효과는 비교적 높을 것으로 나타났다. 또한, 교통사고분석자료를 활용하여 요인분석과 군집분석을 수행한 결과, 충청권의 도시유형별 사고특성은 도시형지역보다는 농촌형지역일수록 교통사고위험도가 높아짐을 확인했으며, 특히 보행자사고 및 농기계사고의 위험도가 높음을 확인했다. 따라서 본 연구에서는 분석결과를 토대로 충청권 교통안전수준의 향상을 위해 보행자 건널목 인식등, 교통약자 노견용 인식등, 농기계 운행 안전 LED전광판 등의 기술을 제안한다.

• 지역연구
• /
• 제11권1호
• /
• pp.21-30
• /
• 1995

This paper deals with the forecasting models for traffic accident by region. Its objectives are to develop the appropriate model for projecting the accident and to analyze the regional characteristics of the accident model. The main results are as follow. First, the literature review, statistical tests and sensitivity analyses show that the joint model combined both PTM and Exponential functions is appropriate to project the traffic accidents by region. Second, the statistical analyses by region. Second, the statistical analyses on the regional accident models indicate that the levels of significance in terms of t-value, $R^2$ and F-value are very high. Finally, the comparative analyses among regions show that the regional differences on the accident patterns can be explained by the joint models and the accident indices (parameters, $P_{max}$, 1/b, $\eta$ etc.) of each region.

• 한국ITS학회 논문지
• /
• 제22권2호
• /
• pp.133-146
• /
• 2023

도심지역의 교통 상태는 효과적인 교통 운영과 교통 제어를 수행하는 데 필수 요소이다. 하지만 교통 상태를 얻기 위해서 수많은 도로 구간에 교통 센서를 설치하는 것은 막대한 비용이 든다. 이를 해결하기 위해서 시장침투율이 높은 센서인 차량 블랙박스 카메라를 이용하여 교통 상태를 추정하는 것이 효과적이다. 하지만 기존의 방법론은 객체 추적 알고리즘이나 광학 흐름과 같이 계산 복잡도가 높고, 연속된 프레임이 있어야 연산을 수행할 수 있다는 단점이 존재한다. 이에 본 연구에서는 심층학습 모델로 차량과 차선을 탐지하고, 차선 사이의 공간을 관심 영역으로 설정하여 해당 영역의 교통밀도를 추정하는 방법을 제안하였다. 이 방법론은 객체 탐지 모델만을 이용해서 연산량이 적고, 연속된 프레임이 아닌 샘플링된 프레임에 대해 교통 상태를 추정할 수 있다는 장점이 있기에, 보유하고 있는 컴퓨팅 자원에 맞는 교통 상태 추정이 가능하다. 또, 도심지역에서 운행하는 서로 다른 특성의 2개의 버스 노선에서 수집한 블랙박스 영상을 검증한 결과, 교통밀도 추정 정확도가 90% 이상인 것을 확인하였다.

• 대기
• /
• 제31권2호
• /
• pp.157-170
• /
• 2021

To investigate the effect of snowfall on the traffic of general roads in Gangwon-do, case analysis was performed in Gangneung, Pyeongchang, and Chuncheon using ASOS (Automated Synoptic Observing System) snowfall data and VDS (Vehicle Detector System) traffic data. First, we analyzed how much the traffic volume and speed decrease in snowfall cases on regional roads compared to non-snow cases, and the characteristics of monthly reduction due to snowfall were investigated. In addition, Pearson correlation analysis and regression analysis were performed to quantitatively grasp the effect of snowfall on traffic volume and speed, and sensitivity tests for snowfall intensity and cumulative snowfall were performed. The results showed that the amount of snowfall caused decrease both in the traffic volume and speed from usual (non-snowfall) condition. However, the trend was different by region: The decrease rate in traffic volume was in the order of Gangneung (17~22%), Chuncheon (14~17%), and Pyeongchang (11~14%). The decrease rate in traffic speed was in the order of Chuncheon (9~10%), Gangneung (8~9%), Pyeongchang (5~6%). No significant results were found in the monthly decrease rate analysis. In all regions, traffic volume and speed showed a negative correlation with snowfall. It was confirmed that the greater the amount of traffic entering the road, the greater the slope of the trend line indicating the change in snowfall due to the traffic volume. As a result of the sensitivity test for snowfall intensity and cumulative snowfall, the snowfall information at intervals of 6-hours was the most significant.

• Journal of Electrical Engineering and information Science
• /
• 제2권5호
• /
• pp.33-39
• /
• 1997

In this paper we proposed a new traffic load shedding scheme which maximizes the throughput of traffic control by decreasing the load of the hot-spot cell using minimum load cell selection (MLCS) algorithm and deployed control flow of calls to define characteristic for hadoff region. we compared the performance of the random shedding approach with that of the proposed algorithm. The results of simulation show that MLCS algorithm minimizes the cal blocking rate under a high-density traffic compared to the random shedding scheme.

• 디지털융복합연구
• /
• 제18권10호
• /
• pp.33-39
• /
• 2020

대형 교통사고는 많은 인명피해를 동반한다. 교통사고를 효율적으로 대처하기 위해선 하루 동안 발생할 수 있는 최대 교통사고 수와 사망자 수, 중상자 수가 정량적으로 제시되어야 한다. 본 연구는 교통사고분석시스템에서 제공하는 2005년부터 2018년까지 전국에서 발생한 일 최대 교통사고 수, 사망자 수, 중상자 수 자료를 사용하여 15년, 30년, 50년에 한 번 발생할 수 있는 최대값을 제시하고자 한다. 지역별 교통사고의 특성을 살펴보기 위해 수도권, 충청권, 경북권, 호남권, 경남권으로 구분하여 일반화극단치분포(GEV분포)에 적합시켰다. GEV분포의 모수는 L-적률추정법으로 추정하였고, Anderson Darling 검정과 Cramer-von Mises 검정으로 분포의 적합성을 확인하였다. 분석결과 50년에 한 번 발생할 수 있는 일 최대 교통사고 수는 수도권 401건, 경남권 168건, 경북권 455건, 충청권 136건, 호남권 205건이다. 인구수와 자동차 등록수가 많은 수도권에 비해 경북권은 면적이 넓고 산지지형이 많으며 산업공단으로 인한 물류이동이 많아 교통사고 수가 상대적으로 높게 나타났다.

• 한국소음진동공학회논문집
• /
• 제17권5호
• /
• pp.398-404
• /
• 2007

In this study, the sensitivity of annoyance was investigated by the subjective jury test for the variations of the frequency components along with various sound pressure levels of sixteen environmental noise sources. Annoyance was, also, evaluated for the road traffic noises. Sound pressure levels were $54{\sim}84\;dB$ which individually divided frequency components with eight bands of equally three bark bands. The results show that vehicle traffic noise is recognized as the most serious environment noise source. The sensitivity of human perception of annoyance in frequency bands is quite different from A-weighting curve. The annoyance found out to be more sensitive in high frequency region and reached its maximum in 3.4 kHz.