• Korean Journal of Remote Sensing
• /
• v.38 no.5_3
• /
• pp.939-951
• /
• 2022

The most important factor in designing autonomous driving systems is to recognize the exact location of the vehicle within the surrounding environment. To date, various sensors and navigation systems have been used for autonomous driving systems; however, all have limitations. Therefore, the need for high-definition (HD) maps that provide high-precision infrastructure information for safe and convenient autonomous driving is increasing. HD maps are drawn using three-dimensional point cloud data acquired through a mobile mapping system (MMS). However, this process requires manual work due to the large numbers of points and drawing layers, increasing the cost and effort associated with HD mapping. The objective of this study was to improve the efficiency of HD mapping by segmenting semantic information in an MMS point cloud into six classes: roads, curbs, sidewalks, medians, lanes, and other elements. Segmentation was performed using various machine learning techniques including random forest (RF), support vector machine (SVM), k-nearest neighbor (KNN), and gradient-boosting machine (GBM), and 11 variables including geometry, color, intensity, and other road design features. MMS point cloud data for a 130-m section of a five-lane road near Minam Station in Busan, were used to evaluate the segmentation models; the average F1 scores of the models were 95.43% for RF, 92.1% for SVM, 91.05% for GBM, and 82.63% for KNN. The RF model showed the best segmentation performance, with F1 scores of 99.3%, 95.5%, 94.5%, 93.5%, and 90.1% for roads, sidewalks, curbs, medians, and lanes, respectively. The variable importance results of the RF model showed high mean decrease accuracy and mean decrease gini for XY dist. and Z dist. variables related to road design, respectively. Thus, variables related to road design contributed significantly to the segmentation of semantic information. The results of this study demonstrate the applicability of segmentation of MMS point cloud data based on machine learning, and will help to reduce the cost and effort associated with HD mapping.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.38 no.1
• /
• pp.81-90
• /
• 2018

The purpose of this study is to evaluate the effect of downward speed limit of urban arterial roads at 29 sites in Busan Metropolitan Police Agency to reduce road traffic accidents from '10 to '15. As a result of analyzing the traffic accidents occurred for 1~3 years after the decrease in the speed limit, the number of traffic accidents decreased by 3.09% and the number of injured persons decreased by 8.76%, but the number of deaths decreased by 36.73% The results of this study are as follows. The average speed reduction rate of 6.31km/h was decreased by investigating the change of the vehicle speed before and after the downward speed limit, and the change of average speed was statistically significant in most of the sections. The rate of compliance with the speed limit increased by 10.26% p, which is considered to have greatly improved overall traffic safety. A survey conducted by residents near the target area with a lower speed limit showed that 57.9% of the respondents felt the driving speed of the vehicle was lowered. However, this project was focused on vehicles with limited speed road signs and traffic safety signs, Only 25.8% of respondents said walking safety was improved. In the future, it is necessary to consider the safety of pedestrians by improving roads around roads such as road curvature and separation. In addition, there is a clear positive result in terms of decreasing the fatal accidents in the downward speed limit zone of Busan Metropolitan Subway. However, more detailed analysis is needed for the 29 accidents. Therefore, it is expected that traffic practitioners will be able to utilize it as a basis to increase the accident reduction effect by setting an appropriate speed limit based on the easy and objective grounds.

• Journal of Korean Society of Transportation
• /
• v.20 no.5
• /
• pp.113-130
• /
• 2002

In this paper, we developed DNL(Dynamic Network Loading) model based on Moving cell theory to analyze the dynamic characteristics of traffic flow in congested network. In this paper vehicles entered into link at same interval would construct one cell, and the cells moved according to Cell following rule. In the past researches relating to DNL model a continuous single link is separated into two sections such as running section and queuing section to describe physical queue so that various dynamic states generated in real link are only simplified by running and queuing state. However, the approach has some difficulties in simulating various dynamic flow characteristics. To overcome these problems, we present Moving cell theory which is developed by combining Car following theory and Lagrangian method mainly using for the analysis of air pollutants dispersion. In Moving cell theory platoons are represented by cells and each cell is processed by Cell following theory. This type of simulation model is firstly presented by Cremer et al(1999). However they did not develop merging and diverging model because their model was applied to basic freeway section. Moreover they set the number of vehicles which can be included in one cell in one interval so this formulation cant apply to signalized intersection in urban network. To solve these difficulties we develop new approach using Moving cell theory and simulate traffic flow dynamics continuously by movement and state transition of the cells. The developed model are played on simple network including merging and diverging section and it shows improved abilities to describe flow dynamics comparing past DNL models.

• Journal of Korean Society of Transportation
• /
• v.23 no.6 s.84
• /
• pp.91-102
• /
• 2005

When it comes to the process of information storage, people are likely to organize individual information into the forms of groups rather than independent attributes, and put them together in their brains. Likewise, in case of finding the shortest path, this study suggests that a Hierarchical Road Network(HRN) model should be selected to browse the most desirable route, since the HRN model takes the process mentioned above into account. Moreover, most of drivers make a decision to select a route from origin to destination by road hierarchy. It says that the drivers feel difference between the link travel tine which was measured by driving and the theoretical link travel time. There is a different solution which has predicted the link travel time to solve this problem. By using this solution, the link travel time is predicted based on link conditions from time to time. The predicated link travel time is used to search the shortest path. Stochastic Process model uses the historical patterns of travel time conditions on links. The HRN model has compared favorably with the conventional shortest path finding model in tern of calculated speeds. Even more, the result of the shortest path using the HRN model has more similar to the survey results which was conducted to the taxi drivers. Taxi drivers have a strong knowledge of road conditions on the road networks and they are more likely to select a shortest path according to the real common sense.

• Journal of Korean Society of Transportation
• /
• v.29 no.3
• /
• pp.31-40
• /
• 2011

Generally, inappropriate driving conditions including geometric, traffic environmental, and driver psychological problems may be critical reasons of traffic accidents. Under this circumstance, various types of facilities have been installed to improve traffic safety by itself or as a set consisting of several other traffic safety facilities. In general, traffic accidents occur by several reasons combined rather than only a single reason, and thus the safety effect of the safety facilities cannot be simply analyzed with only a single improvement. For the study, traffic accident data on the roadway segments of interest are collected along with field survey data. For the analysis, various alternative analysis methods were evaluated in terms of assessing accident reduction from various types of traffic safety improvements. Among the alternative methods tested including simple before-and-after evaluation method, before-and-after evaluation yoked comparison, and Comparison Group (C-G) method, it was found that the C-G method is the most effective method for analyzing the traffic safety improvement effect. Adopting the C-G method, both single and multiple safety improvements were analyzed. The results from this study can potentially be applied to decide the best type of treatments to improve traffic safety as well as to measure the accident reduction effects from the treatments.

• 한국방재학회:학술대회논문집
• /
• 2011.02a
• /
• pp.38-38
• /
• 2011

일반적으로 레일마모는 열차의 주행안전 및 승차감에 미치는 영향이 크고, 소음 진동의 주요원인으로 작용한다. 또한 레일마모가 발생할 경우 궤도구조의 파괴를 촉진시킴으로써 차량 및 궤도유지보수비를 크게 증가시킨다. 따라서 구간 특성 및 환경 영향 인자 등 현장에서 발생하는 마모 원인을 체계적으로 분석함으로써 마모를 저감할 수 있도록 차량운행 조건과 선로선형 및 궤도구조를 설계하는 것은 중요한 과제이다. CART(Classification And Regression Tree; 분류와 회귀나무) 분석은 패키지화된 좋은 분류 및 예측도구 기법으로 나무의 상위 분리수준에서 일반적으로 나타나는 가장 중요한 입력변수들을 사용하는 등의 입력변수를 선정하는 경우 매우 유용하다. 본 연구에서는 다변수 구간특성 및 환경인자를 고려한 검측 자료 상관관계 분석을 위한 회귀 나무기반 모델(TBM: Tree Based Model) 분석 수행을 위해 지하철 2호선 마모 데이터와 마모 데이터에 영향을 미치는 각종 다변수 구간특성 및 환경인자를 사용하였다. 2호선 지하철의 구간특성 인자 및 환경인자는 레일의 종류, 레일의 위치, 도상, 곡률반경, 캔트 슬랙 및 운행 일수 등으로 구분하였다. 레일의 종류는 ks-50kg과 ks-60kg 두 종류의 레일이 있으며, 레일의 위치는 지상과 지하로 크게 구분할 수 있다. 도상은 콘크리트 도상, 자갈 도상과 일부 구간의 방진상 콘크리트 도상으로 구분할 수 있으며, 곡률반경은 직선구간과 완화곡선 구간 및 최소 250m부터 627m까지 분포된 원 곡선 구간으로 구분할 수 있다. 캔트 간격은 최소 96cm 부터 120cm 간격으로 구분하며, 슬랙은 5~9cm에 분포하고, 운행 기간은 해당 기간 동안 유지보수 이력이 없는 구간을 선정하여 2005년부터 2006년까지 4번에 걸쳐 검측된 지하철 2호선 내선 마모데이터를 사용하였다. 총 X1부터 X7까지 총 7개의 구간특성 또는 환경특성을 영향인자로 선정하였으며, 이러한 영향인자에 의해 결정되는 종속 인자로 Y1인 직마모와 Y2인 측마모를 선정하여 이 중 실질적으로 지하철 궤도의 성능 평가에 주요 판단인자로 사용되는 측마모와 구간특성 및 환경영향인자와의 상관관계 분석을 수행하였다. 해당 마모 데이터가 검측되는 기간 동안 유지보수 이력이 없는 12272 point의 데이터를 검출하였고 CART 프로그램을 이용하여 데이터를 분석하였으며, CART 프로그램의 해석을 위해 종속변수인 직마모량은 각 검측 지점의 마모량에 해당하는 등급으로 변환하여 분석을 수행하였다. 레일의 마모에 영향을 미치는 구간특성 및 환경인자와 종속 변수로 사용된 레일의 마모량 사이의 CART를 이용한 상관관계 분석은 실제 구조물에서 영향인자간의 상관 관계와 유사하며, 추후 연구에서는 이를 바탕으로 하여 정량화된 검측 데이터를 종속변수로 하여 구간특성 또는 환경인자 등 외부 영향인자를 고려한 궤도 검측데이터와의 상관관계 분석을 수행할 계획이다.

• International Journal of Highway Engineering
• /
• v.9 no.3
• /
• pp.83-89
• /
• 2007

When allocating traffic signal at the signalized intersection, minimum green time and clearance time for bicyclists should be significantly considered in order to enhance safety aspects to bicyclists when crossing intersections, especially where intersections with exclusive bicycle paths that are physically separated from pedestrians. In this study, field measurements related to bicycle crossing time, including minimum peen time and clearance time, were collected and analyzed according to bicycles crossing types at the signalized intersections where high rate of bicyclists exists. Three types of bicycle crossing are defined as follows 1) stopping: completely stop before crossing (at least one foot on found) 2) riding: crossing with riding bicycle 3) pulling: crossing without riding bicycles. Minimum green time based on pedestrian speeds should be used as crossing time in this case. For bicyclists, speed of bicycle that is applicable to estimate the minimum green time is in the 1.36m/sec($15^{th}$ percentile) to 1.60m/sec($25^{th}$ percentile) range in case of its stopping. Also it is in the 0.75($15^{th}$ percentile) to 0.87($25^{th}$ percentile) range for pulling at crosswalk. In addition, speed of bicycle to consider for calculating the clearance time is in the 2.51m/sec($15^{th}$ percentile) to 2.79m/sec($25^{th}$ percentile). These values also resulted from $15^{th}$ percentile or $25^{th}$ percentile speeds of riding. The results of this study are expected to be supported in traffic signal allocation process, reflecting bicyclists' characteristics.

• Journal of the Institute of Convergence Signal Processing
• /
• v.23 no.1
• /
• pp.21-27
• /
• 2022

Traditional automobile or 2-wheeled robot have limitations on mobility because of their mechanical structure. As traditional automobile is being replaced by electric cars, robot technology is applied to the car industry. In robotics, many researchers worked on omnidirectional mobile robot and produced lots of noticeable results. However in many of the results, specialized wheels such as Mecanum wheels are required. That imposes restrictions on robot speed and outdoor driving. We proposed a 2-wheeled modular robot that has omnidirectional mobility without using specialized wheels. In this paper, we propose a 4-wheeled omnidirectional mobile robot that consists of those two modular robots. The proposed robot adopts electric brakes to combine wheel housings and the robot body or to separate wheel housings from the robot body. Two absolute-type encoders and four incremental encoders are used to control the position of the wheel housing and velocities of the wheels. The proposed robot has omnidirectional mobility and can move fast and outdoor with normal tire wheels. We implemented the proposed robot and the feasibility and stability of the robot is verified by two separate experiments.

• Biomedical Science Letters
• /
• v.2 no.1
• /
• pp.71-90
• /
• 1996

There is considerable current interest in the effect of regular vigorous exercise and in particular endurance-running as a possible measure in improving myocardial function. Some data indicate that the aging heart may actually suffer from vigorous endurance exercise. On the contrary appropriate exercise in aged animals improves myocardial function and aerobic energy metabolism. So far there is relatively little data to indicate that endurance exercise is in fact beneficial in improving myocardial function or damaging to heart of aged animals. The present investigation aimed to study the possible effect of a long range treadmill training program on the heart in aging rats. Male rats aged 3, 10, and 20 months were divided at random into a control (sedentary) and an exercise group. The training group was exercised for 5 days a week on an automated treadmill for 20minutes at 18m/min over a period of 5 months. The exercise regimen of our experiments did not cause any significant changes in the tissues and ultrastructural as com-pared with sedentary age-matched control. Tissues and ultrastructures of myocardial cells in trained group aged 8 months are intact and well organized as well as sedentary control group. Age associated tissue and ultrastructural changes of trained group aged 15 months included : an increase in transformed mitochondria, vacuoles, lysosomes, lipid droplets and early lipofuscin. But the trained heart did not show significant difference in tissue and ultrastructural properties from those of sedentary controls. Endurance-trained group aged 25 months showed significant qualitative tissue and ultrastructural difference as compared with age-matched controls. In addition to those found in 25 months control group, focal necrosis, myofibril fraying, hypercontraction band, seperation of intercalated discs, degenerating nucleus and infiltration of collagenous fiber into myocyte were noted in trained 25 months group. The stereological examination of the mi-crographs disclosed no significant difference in the myoflbril, mitochondrion, sarcotubule and in-terstitium volume density and surface density of mitochondrial cristae and numerical density of mitochondria between trained and control group aged 8 and 15 months. In the trained 25 months group, significant increase in volume density of interstitium, lipofucsin granule were shown as compared to untrained age-matched control. On the other hand, significant decrease in mitochondrion volume density was shown. The myofibril volume density did not differ between trained and control group although trained group showed slight increase. From the data obtained a reduced mitochondria/myofibrils ratio was found in trained rat heart aged 25 months and there was no difference between trained and control rat aged 15 months. But a slight but not significant increase was found in the trained group aged 8 months as compared with same age control group. Such increase in the ratio in young animals is considered to be of great importance to cardiac pumping and adaptability. Whereas such adaptations don't seem to occur in aged heart muscle. This study proposed that repeated endurance exercise do not cause any significant qualitative and quantitative ultrastructural change of heart muscle in young(3months) and adult (10months) suggesting that the heart is able to adapt to the exercise. On the contrary, the repeated endurance exercise stress may actually induce degenerative changes in the aged heart muscle(20months).