• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.01a
• /
• pp.269-270
• /
• 2023

재해, 재난 발생 시에 구조대가 건물 내부나 지하철 등, 특정 구역 내의 대피하지 못한 잔류인원을 제대로 파악하데 어려움을 겪는다. 이를 개선하고자 YOLO와 DeepSORT를 활용하여 통행자를 인식하여 특정 구역의 잔류인원을 파악하고 이를 서버를 통해 확인할 수 있는 시스템을 개발하였다. 실시간 객체인식 알고리즘인 YOLOv4-tiny와 실시간 객체추적기술인 DeepSORT 알고리즘을 이용하여 제안한 방법을 Ubuntu환경에서 구현하고, 실내 상황에 맞춰 통행자 동선을 고려해서 적용하였다. 개발한 시스템은 인식된 통행자 객체방향으로 출입을 구분하여 데이터를 서버에 저장한다. 이에 따라 재해 발생 시 구역의 잔류인원을 파악하여 빠르고 효율적으로 요구조자 위치와 인원을 예측할 수 있다.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.4 no.2 s.7
• /
• pp.13-22
• /
• 2005

The purpose of this research is how to estimate the traffic queue length in the signal intersection accurately. The current traffic queue length algorithm in COSMOS has been using the congestion diagram which comes from the speed of an average separated vehicle - using average vehicle length and the occupancy time from loop detectors. So some errors were occurred by the speed estimation method using average vehicle lengths. And Operators had been difficult to optimize some variables for measuring the traffic queue length estimation algorithm in COSMOS. Therefore the traffic queue length estimation algorithm on the basis of the relation between distances and occupancy rates from loop detectors was developed in this thesis. This thesis had the advantage of using occupancy rates which came out from loop detectors easily and no need to optimize some variables for the established algorithm in COSMOS. And the results of testing this algorithm in some sites which had installed COSMOS system showed better results than COSMOS system's results. But it was noted that further studies which carry it out in various sites and under various cases are necessary for applying to actual intersections.

• Korean Journal of Construction Engineering and Management
• /
• v.12 no.2
• /
• pp.81-88
• /
• 2011

Providers haven't recently had a flexible construction cost estimation system to meet various needs of consumers about public housing. So the subject of this study is to estimate construction cost reasonably in early project stage of public housing and then develop reliable means which is able to support construction cost management and establish a adequate funding investment plan as a provider. In this study, Regression analysis was performed by the case on 20 public apartment complex which were designed from the first half of 2007 to the first half of 2008. A total construction cost of construction, civil engineering, machinery, elevator, land scape, electricity and communication work was used as one sample for increasing explanation and representativeness of the case. In addition, The total construction cost which is devided into design, contract and completion cost was variously analysed for increasing relevance of model and actual utilization. The result of estimation model based on a total construction cost set up completion and design cost showed that error rate is within 2%, which is a excellent result. The estimation model of the construction cost developed by this study is expected to estimate approximate construction cost which is adjacent real construction cost in early stage of the project by using some data.

• Tuberculosis and Respiratory Diseases
• /
• v.39 no.2
• /
• pp.141-149
• /
• 1992

In order to establish predicted normal values of volume of isoflow($V_{iso}\dot{V}$), $V_{iso}\dot{V}$ were measured in a total 234 healthy adults, consisting of 78 male nonsmokers, 108 male smokers and 48 female nonsmokers. The volumes of isoflow were determined through analysis of maximal expiratory flow volume curves recorded with a mixture of 80% helium and 20% oxygen(He-$O_2$ gas) following one(1VC) and three vital capacity(3VC) maneuver and with room air. Based upon the correlation of the observed values of the parameters to age and physical characteristics, the following regression formulars for the prediction of the parameters, using age in year(A) as a variable, were obtained. 1VC method ($V_{iso}\dot{V}_1$) Male smoker : 0.614A+2.347 Male nonsmoker : 0.370A-2.792 Female nonsmoker : 0.588A-2.459 3VC method ($V_{iso}\dot{V}_3$) Male smoker : 0.467A+1.696 Male nonsmoker : 0.288A-1.538 Female nonsmoker : 0.367A-0.114.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.19 no.11
• /
• pp.338-343
• /
• 2018

Recently, a computer numerical control (CNC) machine is used widely for mold making in various industries. In the operation of a CNC machine, the production quality and safety of workers are becoming increasingly important as the product process increases. A variety of tool life prediction studies has been conducted to standardize the quality of production and improve reproducibility. When the tool life is predicted using the conventional tool life equation, there is a large error between the experimental result and result by the conventional tool life equation. In this paper, an algorithm that can predict the precise tool life was implemented using a genetic algorithm.

• Environmental and Resource Economics Review
• /
• v.13 no.4
• /
• pp.685-715
• /
• 2004

This study applies two-choice model to identify the major determinants of NIMBY attitudes when a large-scale composting facility is built around a residential area. Using a survey data of residents in Jeju City, Korea, logit estimation is implemented. The empirical results are consistent with the implication of the specified model: a representative resident's NIMBY attitude is positively (negatively) affected by "Negative Neighborhood Characteristic Variables" ("Positive Wealth Attribute variables"). Socio-demographic variables may be summarized as mostly statistically insignificant, which implies that policy makers may have to take into consideration their region-specific socio-demographic factors instead of simply emulating the policies which have been successful elsewhere.

• Journal of the Korean Society of Safety
• /
• v.13 no.2
• /
• pp.170-180
• /
• 1998

This study presents a method for calculating the concentration of hydrocarbon releases in enclosed areas using empirical equations of evaporation rate. The approach of the method is to estimate the hydrocarbon exposure concentration in the air under conditions assumed. A methodology for assessing risk was suggested to individual risk assessment to exposed workers or others by probit expressions. The toxicity criteria and available human exposure data were examined and guidelines for risk assessment suggested for benzene-air and toluene-air systems. The value of probit constants with mole fractions of lethal concentrations in a mixture of hydrocarbons and a non-toxic substance was predicted. The probit values calculated with mole fractions can be used to estimate guidelines to prevent toxicity within enclosed working areas.

• Journal of Teachers' Pension
• /
• v.1
• /
• pp.157-185
• /
• 2016

본 연구에서는 비정상적 사건을 정의하고 이에 따른 비정상적 위험의 구체적인 유형을 파악하며, 이와 관련된 사학연금의 위험관리 체계에 대한 검토와 함께 비정상적 위험에 효과적으로 대응할 수 있는 자산운용방안을 제시하였다. 우선 비정상적 사건을 '과거 자료를 이용한 발생확률의 추정이나 발생여부에 대한 예측이 불가능하며 따라서 이의 발생 가능성을 사전에 고려하고 대비하는 사전적인 대처가 어려운 사건으로서 자산운용과 위험관리에 무시할 수 없는 영향을 미치는 사건'으로 정의하였으며, 이의 구체적인 형태로서 금융위기를 포함하는 9가지 사건 유형을 파악하였다. 동비정상적 사건들은 포트폴리오 투자를 통한 자산운용에서 개별자산군의 기대수익률과 위험 및 자산군 사이의 상관관계에 영향을 미쳐, 기존의 자산배분안의 최적성을 상실시키고 위험수준의 측정치인 VaR값을 과소 또는 과대추정하게 할 수 있는 것으로 분석되었다. 한편 비정상적 사건의 해외 사례에 대한 분석에서는 비정상적 사건의 영향이 개별 사건마다 다양한 양태로 발현되는 것이 관측되었다. 본 연구에서는 사학연금의 현행 자산배분 체계가 이와 같은 비정상적 사건의 영향에 적절하게 대응하기 어려운 상황이라고 진단하였으며, 비정상적 사건에 적절히 대응하기 위한 자산관리방안의 일환으로서 일별 수익률 자료를 사용한 비정상적 사건의 영향 평가방안을 제시하였다. 한편, 사학연금의 현행 위험관리 체계는 비정상적 사건의 발생에 적절하게 대응할 수 있는 것으로 평가되었다

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.19 no.4
• /
• pp.623-632
• /
• 2018

Predicting the behavior of rheological polymers is highly shear rate- and temperature-dependent. The Cross-WLF viscosity model has become a powerful solution that describes the shear rate- and temperature-dependent characteristics. To estimate the behavior of polymers in computational simulations, the coefficients of the Cross-WLF model should be well identified. An identification technique was proposed to determine the Cross-WLF viscosity model coefficient. The assumption is that the Cross-WLF viscosity model well describes the real characteristics of polymers when the calculated viscosity with the parameters is identical to the reference data. In this study, Auto-desk Moldflow data were used as a reference. The numerical examples showed that the proposed method accurately identifies the Cross-WLF viscosity model coefficients.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.22 no.2
• /
• pp.54-63
• /
• 2023

Pedestrian traffic needs to be accurately quantified to predict effectively pedestrian traffic accidents, however, pedestrian traffic is more difficult to measure than vehicle traffic. In this study, we suggest the time-and cost-effective application of mobile closed-circuit television (CCTV) using a smartphone as an alternative that can collect and analyze real-time data with little. In the present investigation, the pedestrian-vehicle conflict that can develop into an accident was defined as the pedestrian accident exposure. After installing mobile CCTV in 40 sections of Dongseong-ro, Daegu, the pedestrian accident exposure was estimated through negative binomial regression analysis using the collected data. The results of the analysis showed statistically significant changes in the pedestrian accident exposure variables. Based on the present results, a pedestrian accident exposure estimation model was developed which can be used in sections where pedestrian accidents may occur.