• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.07a
• /
• pp.89-92
• /
• 2022

본 논문에서는 딥러닝 모델 기반 보행자의 GPS 경로를 예측하는 시스템을 제안한다. 다양한 경로 예측 방식들 중 본 논문은 GPS 데이터 기반 경로 예측 연구이다. 시계열 데이터인 보행자의 GPS 경로를 학습하여 다음 경로를 예측하도록 하는 딥러닝 모델 기반 연구이다. 본 논문에서는 보행자의 GPS 경로를 딥러닝 모델이 학습할 수 있도록 데이터 구성 방식을 제시하였으며, 예측 범위에 큰 제약이 없는 예측 딥러닝 모델을 제안한다. 본 논문의 딥러닝 모델에 적합한 파라메터들을 제시하였으며, 우수한 예측 성능을 보이는 결과를 제시한다.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.8
• /
• pp.61-68
• /
• 2022

In this paper, we propose a system to predict the GPS trajectory of a pedestrian based on a deep learning model. Pedestrian trajectory prediction is a study that can prevent pedestrian danger and collision situations through notifications, and has an impact on business such as various marketing. In addition, it can be used not only for pedestrians but also for path prediction of unmanned transportation, which is receiving a lot of spotlight. Among various trajectory prediction methods, this paper is a study of trajectory prediction using GPS data. It is a deep learning model-based study that predicts the next route by learning the GPS trajectory of pedestrians, which is time series data. In this paper, we presented a data set construction method that allows the deep learning model to learn the GPS route of pedestrians, and proposes a trajectory prediction deep learning model that does not have large restrictions on the prediction range. The parameters suitable for the trajectory prediction deep learning model of this study are presented, and the model's test performance are presented.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2016.10a
• /
• pp.810-813
• /
• 2016

기존 IoT 스마트 가로등은 모션 감지 센서를 이용하여 보행자를 감지하고, 가로등의 밝기를 조정하는 형태로 구성된다. 하지만 이러한 방법은 보행자가 나아갈 길을 미리 예측하여 밝혀주지 않는다. 특히 기존 방법은 보행자 현재 위치만 밝힐 뿐, 나아갈 길은 어두운 상태이기 때문에 통행에 불편함을 겪고 있다. 본 논문에서는 보행자 경로를 미리 파악하여 가로등 밝기를 조절하는 방식을 소개한다. 보행자의 현재 위치를 파악하기 위해 모션 감지 센서를 이용하며, 보행자 경로 예측은 모션 벡터를 사용하여 가로등 밝기를 조절한다. 이러한 개선을 통하여 보행자의 편의 증대와 범죄 예방 등 긍정적인 효과를 기대 할 수 있다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.9
• /
• pp.1244-1253
• /
• 2021

In this paper, we propose a method for predicting in advance whether pedestrians will enter the hazardous area after the current time using the pedestrian trajectory prediction method and an efficient simplification method of the trajectory prediction network. In addition, we propose a method to apply KD(Knowledge Distillation) to a small network for real-time operation in an embedded environment. Using the correlation between predicted future paths and hazard zones, we determined whether to enter or not, and applied efficient KD when learning small networks to minimize performance degradation. Experimentally, it was confirmed that the model applied with the simplification method proposed improved the speed by 37.49% compared to the existing model, but led to a slight decrease in accuracy. As a result of learning a small network with an initial accuracy of 91.43% using KD, It was confirmed that it has improved accuracy of 94.76%.

• Fire Science and Engineering
• /
• v.2 no.3
• /
• pp.84-95
• /
• 1988

이 논문에서는 Predtecheuskii와 Milinski에 의해 개발된 보행자 이동을 대비한 설계방법에 대하여 검토하고, 고층사무소 건물에서의 실제 피난훈련에 대하여 기술에서의 실제 피난훈련에 대하여 기술하면서 고충건물에서의 계단을 중심으로 한 출구의 모델을 제공한다. 또한 이 논문은 예측되는 피난 수단과 그것에 대한 규제의 필요성을 간략하게 비교하였다.

• Journal of Navigation and Port Research
• /
• v.39 no.1
• /
• pp.29-35
• /
• 2015

As the numbers of users of domestic passenger ships increased up to 14.5 millions at 2012, the numbers of ships accidents also increased as 2 times than ever. It will be very important to develop technologies related with safety design for onboard passengers and disclose the potential problems. This study performed consciousness survey on ordinary peoples' way finding who have not got any regular anti-disaster training, to develop evacuation models for evacuation feasibility studies. Followings are the results answered by 83 participants for 33 way finding questions. Respondents selected right ways more than 6~18% for 2 ways like T type, U type, Y type passages. But when there are some walkers and/or runners, respondents preferred to select the way where walkers or runners are. And more over the ratio of the ways that runners are on is comparatively higher than walkers. On 'ㅏ'type, 'ㅓ'type and 3 way type passages, even though the walkers and/or runners are affected to answerers, straight way were most preferred. And it is clear that peoples like bright passages. On the other hands, peoples responded as they like right, downward and near stairs more than left, upward and far stairs, respectively. and very few selected escalator and elevator for as evacuation stairs.

• Journal of Korean Society of Transportation
• /
• v.36 no.2
• /
• pp.112-128
• /
• 2018

We are implementing pedestrian priority zone policy to certain districts to reduce greenhouse gas and to develop eco-friendly city which has more focus on pedestrians' walking environment. This policy has contributed to citizens' satisfaction level with improved public transportation service as well as more spacious streets for walk. Despite highly positive influence of pedestrian priority zone policy to the walking environment, we need to anticipate the impact of this to traffic environment as it may have bad effect to the overall traffic flow around the zone where the policy is implemented. This research has analyzed the change of characteristics of traffic accidents around the eco-traffic area of Hang-Gung dong, Suwon city, to understand impact of the pedestrian priority zone policy to the traffic surroundings, with pre-post analysis methodology. As a result, number of accidents related to pedestrians showed decrease as pedestrian priority zone is designed operated with focus to pedestrians. But accidents related illegal U-turn and violation of the traffic signal showed (significant) increase as there was a restriction of turns and decrease of overall traffic speed. To prevent the accidents above, we need to notice drivers to pay special attention before the pedestrian priority zone event, and information from this research should be given to the drivers through safety signs and mobile application at the place near to the event.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2024.05a
• /
• pp.633-634
• /
• 2024

본 연구는 LSTM 모델이 수용 가능한 익명 보행자의 GPS 경로 범위와 훈련 데이터 셋의 크기에 대한 양상 분석을 목적으로 한다. 시계열 데이터인 GPS 경로 그리고 순환 신경망 LSTM 과 입력 구조를 이해하고, 두 가지 실험을 설계하여 LSTM 의 훈련 데이터 셋 수용을 파악한다. 실험에서는 장거리 데이터 셋을 학습한 모델과 그렇지 않은 모델을 비교하고, 훈련 데이터 셋 크기에 따른 학습 모델의 예측 값을 비교한다. 두 실험을 통해 GPS 경로 범위와 학습 가능한 경로의 가짓수에 대한 비교 분석 결과를 제시한다.

• The Journal of Bigdata
• /
• v.8 no.2
• /
• pp.37-44
• /
• 2023

Despite policy and institutional changes to prevent accidents in the workplace, industrial accidents continue to occur. Accidents related to safety at industrial sites are complex tasks that involve considering various variables, and research to address and minimize these accidents is ongoing. In this paper, we studied the trajectory prediction of transporters used in industrial areas. In a transporter work environment measuring 3.4km by 2.3km, the coordinates of transporters were learned through PECNet, resulting in an average error of ADE 1.27m and FDE 1.13m. This research will contribute to preventing and avoiding accidents by predicting not only the path of transporters but also the paths of mobile vehicles and pedestrians in various fields.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.10
• /
• pp.415-420
• /
• 2020

Crowd flows occur in metropolitan railway transit stations, terminals, multiple buildings, and stadiums and are important in ensuring the safety as well as smooth flow of pedestrians in these facilities. In this study, the author developed a new computational analysis method for crowd flow dynamics and applied it to models of transit connecting paths. Using the analysis method, the potential value of the exit was assigned the smallest value, and the potential value of the surrounding grids gradually increased to form the overall potential map. A pathline map was then constructed by determining the direction vector from the grid with large potential value to the grid and small potential. These pathlines indicate basic routes of passenger flow. In all models of the analysis object, the pedestrians did not move to the first predicted shortest path but instead moved using alternative paths that changed depending on the situation. Even in bottlenecks in which pedestrians in both directions encountered each other, walking became much smoother if the entry time difference was dispersed. The results of the analysis show that a method for reducing congestion could be developed through software analysis such as passenger flow analysis without requiring hardware improvement work at the railway station.