• Journal of Digital Contents Society
• /
• v.18 no.1
• /
• pp.109-114
• /
• 2017

In our modern society, the utilization of the advanced drone which is capable of performing variety of tasks has been gradually increasing. In this paper, we present an application, similar to the prototype "Skycall" that had been introduced in the MIT Senseable City. To assess this concept, we have implemented a prototype of drone-based pedestrian navigation depending on the Android smartphone. Our system is not only able to guide the user in a very complicated place, where buildings are compacted, but also to block unauthorized visitors from accessing the facilities. And we discuss some problems we found and suggest the direction to address them.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.39 no.2
• /
• pp.21-34
• /
• 2002

Different from ordinary mobile robots used in a well-structured industrial workspace, a guide mobile robot for the visually impaired should be designed in consideration of a moving obstacle, which mostly refers to pedestrians in intentional motions. Thus, the navigation of the guide robot can be facilitated if the intention of each detected obstacle can be known in advance. In this paper, we propose an inference method to understand an intention of a detected obstacle. In order to represent the environment with ultrasonic sensors, the fuzzy grid-type map is first constructed. Then, we detect the obstacle and infer the intention for collision avoidance with the CLA(Centroid of Largest Area) point of the fuzzy grid-type map. To verify the proposed method, some experiments are performed.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2016.10a
• /
• pp.506-509
• /
• 2016

We present real-time pedestrian detection that exploit accuracy of Faster R-CNN network. Faster R-CNN has shown to success at PASCAL VOC multi-object detection tasks, and their ability to operate on raw pixel input without the need to design special features is very engaging. Therefore, in this work we apply and adjust Faster R-CNN to single object detection, which is pedestrian detection. The drawback of Faster R-CNN is its failure when object size is small. Previously, small sized object problem was solved by Scale-aware Network. We incorporate Scale-aware Network to Faster R-CNN. This made our method Scale-aware Faster R-CNN (DF R-CNN) that is both fast and very accurate. We separated Faster R-CNN networks into two sub-network, that is one for large-size objects and another one for small-size objects. The resulting approach achieves a 28.3% average miss rate on the Caltech Pedestrian detection benchmark, which is competitive with the other best reported results.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2013.10a
• /
• pp.862-864
• /
• 2013

Recently unstable weather condition has been going on then the damage by heavy rains, typhoons and stream flooding has been increasing continuously. In particular, the urban damage by flooding of rivers or roads has been occurred frequently. It could block all the transportation facilities within cities for restricting traffic. In this study, we suggest the method of utilization of the information about event and accident on roads through the real-time traffic information to cope with dangerous situation of roads.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2015.10a
• /
• pp.1384-1387
• /
• 2015

고령화 사회로 진입하면서 노인들은 노화과정에 의한 보행능력의 감소 및 근력 약화와 같은 신체적 변화로 인해 잦은 낙상을 경험한다. 이에 따라 낙상 사고를 감지하는 연구가 활발히 진행되고 있다. 낙상은 사전 예방도 중요하지만 사고 발생 후의 신속한 대처도 중요하다. 낙상을 감지하고 의료진에게 즉시 낙상정보를 제공하여 후속적 조치를 취하는 것은 사고 후 대처의 핵심이다. 본 논문에서는 스마트폰 환경에서 사용자의 낙상 후 방향을 판별하기 위해 두 가지 센서 데이터의 특정 값들을 추출하였으며, 이에 5 가지 기계학습 알고리즘을 적용하였다. 사용자는 스마트폰을 착용한 상태로 전후좌우 4 방향 낙상 실험을 진행하며 스마트폰 내에 내장된 3 축 가속도 센서와 3 축 자이로 센서값을 측정한다. 피험자 11 명을 대상으로 낙상 실험 결과, 5 가지의 분류기 중 k-NN에서 98.6%의 인식률을 나타내었다. 뽑아낸 특징 값과 분류 알고리즘은 낙상의 방향 검출에 유용한 것으로 판단된다.

• Science of Emotion and Sensibility
• /
• v.23 no.4
• /
• pp.105-116
• /
• 2020

The purpose of this study is to progress the limitations and disadvantages of existing safety clothing by applying high technology to current safety clothing that is produced and distributed only with fluorescent fabrics and retroreflective materials. Therefore, the industrial suspender-type safety belt and engineering technology are introduced, designed, and fabricated to help save a life in an emergency. First, the suspender-type safety belt to be developed is designed to emit light by LED attached to the film, and the body of the belt-wearer is recognized from a distance through retroreflection from the flashing LED. It aims to support people's safety by preventing accidents during roadside work, rescue activities, and sports activities at night. Second, with the development of advanced devices when the user is in an unconscious state due to distress or falls into an unconscious state due to distress or accident, the tilt sensor of the control unit attached to the belt automatically detects the angle of the human body and generates light and sound. It is intended to further enhance the utilization by mounting a sensing and signaling device that generates a distress signal and shaping it in the form of a belt attached to a vest that can be easily detached from the outside of the garment. When the wearer falls due to an accident, the tilt sensor of this belt detects the angle change and then the controller generates a high-frequency sound and repeated LED blinking signals at the same time. In the case of conventional safety vests, it is almost impossible to detect that the person is wearing a vest when there is no ambient light, but in case of the safety belts in this study, the sound and light signals of the safety belt enable us to find the wearer within 100 meters even when there is no ambient light.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.19 no.6
• /
• pp.915-936
• /
• 2017

In this study, current status of Korean hazard mitigation guideline for tunnel operation is summarized. It shows that requirement for CCTV installation has been gradually stricted and needs for tunnel incident detection system in conjunction with the CCTV in tunnels have been highly increased. Despite of this, it is noticed that mathematical algorithm based incident detection system, which are commonly applied in current tunnel operation, show very low detectable rates by less than 50%. The putative major reasons seem to be (1) very weak intensity of illumination (2) dust in tunnel (3) low installation height of CCTV to about 3.5 m, etc. Therefore, an attempt in this study is made to develop an deep-learning based tunnel incident detection system, which is relatively insensitive to very poor visibility conditions. Its theoretical background is given and validating investigation are undertaken focused on the moving vehicles and person out of vehicle in tunnel, which are the official major objects to be detected. Two scenarios are set up: (1) training and prediction in the same tunnel (2) training in a tunnel and prediction in the other tunnel. From the both cases, targeted object detection in prediction mode are achieved to detectable rate to higher than 80% in case of similar time period between training and prediction but it shows a bit low detectable rate to 40% when the prediction times are far from the training time without further training taking place. However, it is believed that the AI based system would be enhanced in its predictability automatically as further training are followed with accumulated CCTV BigData without any revision or calibration of the incident detection system.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.18 no.1
• /
• pp.14-26
• /
• 2019

For bus drivers' safe driving, a policy that analyzes the causes of the drivers' traffic accidents and then assists their safe driving is required. Therefore, the Ministry of Land, Infrastructure and Transport set up its plan to gradually expand the equipping of commercial vehicles with FCWS (Forward Collision Warning System) and LDWS(Lane Departure Warning System), from the driver-supporting ADAS(Advanced Driver Assistance Systems). However, there is not much basic research on the analysis of bus drivers' traffic accidents in Korea. As such, the time is appropriate to research what is the most necessary ADAS for bus drivers going forward to prevent bus accidents. The purpose of this research is to analyze how serious the accidents were in the different bus routes and whether the accidents were repetitive, and to give recommendations on how to support ADAS for buses, as an improvement. A model of ordered logit was used to analyze how serious the accidents were and as a result, vehicle to pedestrian accidents which directly affected individuals were statistically significant in all of the models, and violations of regulations, such as speeding, traffic signal violation and violation of safeguards for passengers, were indicated in common in several models. Therefore, the pedestrian-sensor system and automatic emergency control device for pedestrian should be installed to reduce bus accidents directly affecting persons in the future, and education for drivers and ADAS are to be offered to reduce the violations of regulations.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.17 no.5
• /
• pp.173-187
• /
• 2018

This paper proposes a computer vision and deep learning-based technique for surveillance camera system for vehicle counting as one part of parking lot management system. We applied the You Only Look Once version 2 (YOLOv2) detector and come up with a deep convolutional neural network (CNN) based on YOLOv2 with a different architecture and two models. The effectiveness of the proposed architecture is illustrated using a publicly available Udacity's self-driving-car datasets. After training and testing, our proposed architecture with new models is able to obtain 64.30% mean average precision which is a better performance compare to the original architecture (YOLOv2) that achieved only 47.89% mean average precision on the detection of car, truck, and pedestrian.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.12
• /
• pp.1723-1728
• /
• 2021

C-ITS is an intelligent transportation system that can improve transportation convenience and traffic safety by collecting, managing, and providing traffic information between components such as vehicles, road infrastructure, drivers, and pedestrians. In Korea, road infrastructure is being built across the country through the C-ITS project, and various services such as real-time traffic information provision and bus operation management are provided. However, the current state-of-the-art road infrastructure and information linkage system are insufficient to build C-ITS. In this paper, considering the continuity of time in various spatial aspects, we proposed a group-type network-based traffic situation recognition system that can recognize traffic flows and unexpected accidents through information linkage between traffic infrastructures. It is expected that the proposed system can primarily respond to accident detection and warning in the field, and can be utilized as more diverse traffic information services through information linkage with other systems.