• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.13 no.4
• /
• pp.1-11
• /
• 2014

Traffic congestion cost is more likely to occur in the inner city than interregional road, and it accounts for about 63.39% of the whole. Therefore, it is important to mitigate traffic congestion of the inner city. Traffic congestion in the urban could be divided into Recurrent congestion and Non-recurrent congestion. Quick and accurate detection of Non-recurrent congestion is also important in order to relieve traffic congestion. The existing studies about incident detection have been variously conducted, however it was limited to Uninterrupted Traffic Flow Facilities such as freeway. Moreover study of incident detection on the interrupted Traffic Flow Facilities is still inadequate due to complex geometric structure such as traffic signals and intersections. Therefore, in this study, incident detection model was constructed using by Artificial Neural Network to aim at urban arterial road that is interrupted traffic flow facility. In the result of the reliability assessment, the detection rate were 46.15% and false alarm rate were 25.00%. These results have a meaning as a result of the initial study aimed at interrupted traffic flow. Furthermore, it demonstrates the possibility that Non-recurrent congestion can be detected by using car navigation data such as car navigator system device.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.19 no.6
• /
• pp.96-106
• /
• 2020

To solve the limitation of traffic monitoring that occur from embedded sensor such as loop and piezo sensors, the thermal imaging camera was installed on the roadside. As the length of Class 1(passenger car) is getting longer, it is becoming difficult to classify from Class 3(2-axle truck) by using an embedded sensor. The collected images were labeled to generate training data. A total of 17,536 vehicle images (640x480 pixels) training data were produced. CNN (Convolutional Neural Network) was used to achieve vehicle classification based on thermal image. Based on the limited data volume and quality, a classification accuracy of 97.7% was achieved. It shows the possibility of traffic monitoring system based on AI. If more learning data is collected in the future, 12-class classification will be possible. Also, AI-based traffic monitoring will be able to classify not only 12-class, but also new various class such as eco-friendly vehicles, vehicle in violation, motorcycles, etc. Which can be used as statistical data for national policy, research, and industry.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.39 no.6
• /
• pp.609-618
• /
• 2021

The spread and damage of COVID-19 are putting significant pressure on the world, including Korea. Most countries place restrictions on movement and gathering to minimize contact between citizens and these policies have brought new changes to social patterns. This study generated traffic volume data on the scale of a road network using taxi movement data collected in the early stages of the COVID-19 third pandemic to analyze the impact of COVID-19 on movement patterns. After that, correlation analysis was performed with the data of confirmed cases in Daegu Metropolitan City and Local Moran's I was applied to analyze the effect of spatial characteristics. As a result, in terms of the overall road network, the number of confirmed cases showed a negative correlation with taxi driving and at least -0.615. It was confirmed that citizens' movement anxiety was reflected as the number of confirmed cases increased. The commercial and industrial areas in the center of the city confirmed the cold spot with a negative correlation and low-low local Mona's I. However, the road network around medical institutions such as hospitals and spaces with spatial characteristics such as residential complexes was high-high. In the future, this analysis could be used for preventive measures for policymakers due to COVID-19.

• Journal of the Society of Disaster Information
• /
• v.19 no.2
• /
• pp.334-343
• /
• 2023

Purpose: In this study, the cavity data found through ground cavity exploration was combined with underground facilities to derive a correlation, and the ground subsidence prediction map was verified based on the AI algorithm. Method: The study was conducted in three stages. The stage of data investigation and big data collection related to risk assessment. Data pre-processing steps for AI analysis. And it is the step of verifying the ground subsidence risk prediction map using the AI algorithm. Result: By analyzing the ground subsidence risk prediction map prepared, it was possible to confirm the distribution of risk grades in three stages of emergency, priority, and general for Busanjin-gu and Saha-gu. In addition, by arranging the predicted ground subsidence risk ratings for each section of the road route, it was confirmed that 3 out of 61 sections in Busanjin-gu and 7 out of 68 sections in Sahagu included roads with emergency ratings. Conclusion: Based on the verified ground subsidence risk prediction map, it is possible to provide citizens with a safe road environment by setting the exploration section according to the risk level and conducting investigation.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.31 no.4
• /
• pp.269-275
• /
• 2013

Recently, importance of pedestrian road network is getting emphasized as it is possible to provide mobile device users with both route guidance services and surrounding spatial information. However, it costs a tremendous amount of budget for generating and renovating pedestrian road network nationally, which hinder further advances of these services. Hence, algorithms extracting pedestrian road network automatically based on raster data are needed. On the other hand, road dataset generated from raster data usually has unnecessary vertices which lead to maintenance disutility such as excessive turns and increase in data memory. Therefore, this study proposed a method of selecting a proper threshold automatically for separate road entity using not only Douglas-Peucker algorithm but also road attribute data of digital map in order to remove redundant vertices, which maximizes line simplification efficiency and minimizes distortion of shape of roads simultaneously. As a result of the test, proposed method was suitable for automatic line simplification in terms of reduction ratio of vertices and accuracy of position.

• Proceedings of the Korea Contents Association Conference
• /
• 2010.05a
• /
• pp.55-57
• /
• 2010

수많은 이동 노드가 존재하는 유비쿼터스 환경에서 위치 기반 서비스가 중요한 응용 분야로 부상하고 있다. 효율적인 위치 기반 서비스를 제공하기 위해 브로드캐스팅을 이용한 다양한 기법들이 연구 되었지만, 대부분 효율적인 인덱스 구축에 대한 연구이고, 브로드캐스팅 데이터의 크기를 줄이기 위한 기법은 고려되지 않았다. 이에 본 논문에서는 최근 많은 연구가 이루어지고 있는 센서 네트워크와 브로드캐스팅 기법을 활용하여, 객체의 이동 패턴을 고려한 데이터 분산 브로드캐스팅 기법을 제안한다. 제안하는 기법을 수행하기 위한 기반 인프라를 구축하기 위해 도로 네트워크 기반의 센서 클러스터링 기법을 제안하고, 센서 노드에 의해 측정 된 객체의 이동 정보를 기반으로 한 최적의 데이터 분산 브로드캐스팅 기법을 적용한다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2017.01a
• /
• pp.271-272
• /
• 2017

본 논문에서는 교통사고 데이터베이스 구축을 통해 교통사교 현황과 사고 당시의 여러 정황들을 파악할 수 있는 정보를 제공한다. 이 정보들에는 사고 당시의 기상상태, 도로형태, 차종, 연령, 성별 등의 데이터들이 포함되고 이러한 정보들을 바탕으로 데이터베이스 사용자들은 각 사고 별 종합적인 정보를 얻을 수 있다. 이를 통해 정부 당국 외에 보험사 등에 교통사고 관련 정책을 위한 유용한 정보들을 제공할 수 있다. 또한 운전자 개인들에게도 정보들을 제공해 교통사고를 보다 효율적으로 예방할 수 있다.

• Proceedings of the KIEE Conference
• /
• 2002.11c
• /
• pp.353-356
• /
• 2002

본 논문에서는 지능형 도로교통 제어기에 대하여 논한다. 제어기는 Main CPU module, Field I/O module, Display module, communication module, Mother board module로 구성되었다. 각 모듈은 하드웨어의 특성에 따라 분리되어 설계되었고, mother board를 통하여 module 간 데이터를 교환한다 Main CPU module은 입력된 교통 데이터의 처리, Field I/O module은 외부로의 데이터 입출력, Display module은 제어기와 사용자와의 인터페이스, communication module은 제어기의 debugging을 담당한다. 본 논문에서는 하드웨어를 Module화함으로써 필요한 하드웨어의 장/탈착이 용이하고, 제어기를 범용으로 사용할 수 있는 장점이 있다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.11a
• /
• pp.91-94
• /
• 2022

자율주행 자동차 개발 연구가 활발히 진행됨에 따라 객체 검출기의 성능이 중요하게 되었다. 딥러닝 기술의 발전하면서 객체 검출기의 성능도 큰 발전을 이루었다. 그에 따라 도로 위 차량 검출기의 성능도 발전하고 있으나 평상시 낮 도로상황에서 잘 동작하던 모델은 안개가 끼거나 밤 상황이 되면 제대로 동작하지 못하는 문제를 가지고 있다. 이유는 딥러닝 모델이 학습할 때 사용한 데이터셋의 정보에 따라 특정 도메인에 편향된 특성을 학습하기 때문이다. 따라서, 본 논문에서는 객체 검출 신경망에 도메인 판별기를 적용하여 이와 같은 도메인 이동 문제를 극복하는 모델을 제안한다. 모델의 성능을 Cityscapes 데이터셋과 Foggy Cityscapes 데이터셋을 사용하여 평가한 결과, 기존의 특정 도메인에서 학습한 모델보다 제안하는 모델의 검출 성능이 개선된다는 것을 확인하였다.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.16 no.1
• /
• pp.131-146
• /
• 2017

Unlike in the past based on autonomous vehicle development sensor, In recent years, research has been conducted using external data such as LDM to compensate for the disadvantages of sensors. So, this study suggested the construction method of static map that provides road information for autonomous driving of vehicles as LDM - based information. In other to suggest, after review LDM's ISO 14296 and data specification and map of precise roda map of NGII, we had confirmed the correspondence wiht the international standard of NGII specification. As a result of the review, it is relatively good in terms of provided data and information, but the road structure expression is partially incompatible with the international standard. AS it is necessary to supplement about currently specification and method and suggest, this study had suggested ways to supplement the insuficient and to express the road structure.