• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.19 no.4
• /
• pp.617-622
• /
• 2018

In industries such as automobiles and semiconductors, a lot of components are produced using PLCs based on an automatic production system. Therefore, an educational platform is needed to provide training in the use of PLCs in the industrial sites. Conventional educational systems, in which PLCs are employed to control sensors and actuators, have been continuously developed. However, these systems present their circuits only in 2D on the screen, which makes it difficult to understand them during the learning process. To overcome these disadvantages, we propose an educational V-Factory system capable of providing PLC training using virtual reality. In addition, thousands of sensors and actuators can be controlled by the V-Factory system through the proposed virtual I/O driver. The motion of the automatic production system can be visualized using virtual reality.

• Journal of the Korea Computer Graphics Society
• /
• v.24 no.5
• /
• pp.41-48
• /
• 2018

The most natural way to increase immersion and provide free interaction in a virtual environment is to provide a gesture interface using the user's hand. However, most studies about hand gesture recognition require specialized sensors or equipment, or show low recognition rates. This paper proposes a three-dimensional DenseNet Convolutional Neural Network that enables recognition of hand gestures with no sensors or equipment other than an RGB camera for hand gesture input and introduces a virtual reality game based on it. Experimental results on 4 static hand gestures and 6 dynamic hand gestures showed that they could be used as real-time user interfaces for virtual reality games with an average recognition rate of 94.2% at 50ms. Results of this research can be used as a hand gesture interface not only for games but also for education, medicine, and shopping.

• Journal of Broadcast Engineering
• /
• v.28 no.3
• /
• pp.329-332
• /
• 2023

3D Object Detection using deep neural network has been developed a lot for obstacle detection in autonomous vehicles because it can recognize not only the class of target object but also the distance from the object. But in the case of 3D Object Detection models, the detection performance for distant objects is lower than that for nearby objects, which is a critical issue for autonomous vehicles. In this paper, we introduce a technique that increases the performance of 3D object detection models, particularly in recognizing distant objects, by generating virtual 3D vehicle data and adding it to the dataset used for model training. We used a spherical point tracing method that leverages the characteristics of 3D LiDAR sensor data to create virtual vehicles that closely resemble real ones, and we demonstrated the validity of the virtual data by using it to improve recognition performance for objects at all distances in model training.

• The KIPS Transactions:PartC
• /
• v.16C no.4
• /
• pp.505-510
• /
• 2009

Sensor networks are composed of a great number of sensor nodes. Since all sensor nodes are energy-restricted and hard to recharge, it is very important.However, the energy consumption may significantly increase if mobile sources or sinks exist in sensor networks. The reason is that the routing information for mobile sources and sinks needs to be update frequently for efficientdata delivery. The routing algorithm supporting mobile sinks should consider not only continuous data delivery but also the energy consumption of sensor nodes. However, most of the existing research focuses on even energy consumption while the mobility of sinks and sources is rarely consider. In this paper, we propose an efficient routing protocol with multiple virtual grids to reduce energy consumption and improve packets delivery efficiency. Then this paper considers the mobility. Simulation results show that our algorithm can guarantee high data delivery ratio and lower average delivery delay, while consuming lower energy than existing routing protocols in sensor networks.

• Journal of KIISE:Computing Practices and Letters
• /
• v.14 no.4
• /
• pp.341-351
• /
• 2008

Communication in wireless sensor networks comprising a plurality of sensor nodes located in an ad hoc environment is unidirectional in that data gathered by sensor nodes are transmitted to a sink node and not vice versa. In those networks, it is not possible for a server or a gateway to send commands to the sensor nodes to determine whether some previously received data are valid when the data indicate unusual conditions, which makes it difficult to make appropriate reactions to the unusual situations. This paper proposes and implements a TDMA-based sensor network communication protocol named BiWSLP(Bidirectional Wireless Sensor Line Protocol) supporting bidirectional communication capability. The BiWSLP is an extension of the WSLP, a unidirectional sensor network communication protocol based on the TDMA protocol. To test the feasibility of the proposed BiWSLP, we construct a virtual bridge management system capable of sending commands to sensor nodes as well as collecting data from the sensor nodes. Based on the test results of the virtual bridge management system, we show the applicability and advantages of the BiWSLP in terms of energy efficiency and bidirectional communication capability.

• Journal of IKEEE
• /
• v.16 no.2
• /
• pp.86-94
• /
• 2012

Synchronization MAC Protocol such as S-MAC and T-MAC utilize duty cycling technique which peroidically operate wake-up and sleep state for reducing energy consumption. But synchronization MAC showed low energy efficiency because of additional control packets. For better energy consumption, Asychronization MAC protocols are suggested. For example, B-MAC, and X-MAC protocol adopt Low Power Listening (LPL) technique with CSMA algorithm. All nodes in these protocols joining a network with independent duty cycle schedules without additional synchronization control packets. For this reason, asynchronous MAC protocol improve energy efficiency. In this study, a low-power MAC protocol which is based on X-MAC protocol for wireless sensor network is proposed for better energy efficiency. For this protocol, we suggest preamble numbering, and virtual-synchronization technique between sender and receive node. Using TelosB mote for evaluate energy efficiency.

• Proceedings of the Korean Society for Emotion and Sensibility Conference
• /
• 2009.11a
• /
• pp.247-250
• /
• 2009

본 연구는 가상 캐릭터인 3D Pooh 와의 감성 교감이 이루어지는 유아용 감성 교육 게임 개발이 목적이다. Web 카메라를 이용하여 사용자의 위치 정보 및 영역을 검출하고 사용자는 화면에 보이는 가상 공간에 손과 마커를 이용하여 행동을 가하게 된다. 사용자의 행동 입력 방식에 따라 Pooh 의 상태를 나타내는 게이지와 행동의 다양한 변화를 제공하며, 이로써 사용자들이 가상 캐릭터인 3D Pooh 와 인터렉션을 할 수 있도록 제공한다. 기존의 감성 교감 연구 방법 분석 결과 뇌파와 체온 등의 생리신호를 이용한 방법과 사용자의 행동패턴을 분석하는 방법 등이 있으며, 현재는 생리신호를 이용하는 방법이 더 많이 활용되고 있다. 생리신호 측정 시 사용자는 신체에 센서를 착용해야 하므로 착용 부담감과 움직임이 자유롭지 못하다는 단점이 있다. 본 프로그램은 기존 연구의 단점인 센서 착용의 부담감을 줄이고 자유롭고 편리한 인터렉션이 가능하도록 개발하였으며, 직관적인 인터페이스를 기반으로 유아들도 쉽고 편리하게 즐길 수 있도록 구현된 감성 교육 게임이 될 것으로 기대된다. 또한 카메라 이외의 다른 하드웨어 장치 사용을 최소화하여 경제성과 사용 편의성을 동시에 고려한 효율적인 유아용 감성 교육 게임으로 활용될 수 있다.

• The Journal of the Korea Contents Association
• /
• v.8 no.1
• /
• pp.308-317
• /
• 2008

The objects real time rendered in the 3D cyber space can interact with each others according to the events which are happened when satisfying some conditions. But to representing the behaviors with these interactions, too many event conditions are considered because each behavior pattern and event must be corresponded in a one-to-one ratio. It leads to problems which increase the system complexity. So, in this paper, we try to physical method based on elasticity force for representing more realistic behaviors of AI fish and apply to the deformable multi-detection sensor, so we suggest the new method which can create the various behavior patterns responding to one evasion event.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2010.04a
• /
• pp.158-161
• /
• 2010

본 논문에서는 USN 환경에서 수질 모니터링 시스템을 구성하는 데 있어 가장 중요한 요소 중 하나인 센서 매니지먼트의 기능에 대해 소개한다. 센서 매니지먼트는 센서 노드에서의 효율적인 배터리 관리와 전송 데이터의 축약을 위해 센싱 타임과 전송 횟수 등을 관리해주는 기능을 담당한다. 제안된 센서 매니지먼트 시스템은 센싱 횟수를 조절하여 유동적이고 지능적인 센싱을 하면서도, 불필요한 센싱과 데이터 전송 횟수를 줄여서 센서노드의 배터리 수명을 연장시킬 수 있다. 설계한 기능에 대한 평가를 위해 가상의 시나리오 데이터를 구성하여 실험하였으며, 그 결과 일반적인 동일 시간 간격의 센싱 및 전송에 비해 매우 효과적인 방법임을 확인할 수 있었다.

• Proceedings of the KIEE Conference
• /
• 2008.04a
• /
• pp.43-44
• /
• 2008

본 논문에서는 Weighted Cenrtoid Localization(WCL) 알고리즘을 적용하여, 고정 노드 (Anchors) 사이의 측정된 수신 신호 강도(RSSI)로 Zigbee 실내 위치 인식 시스템을 구현하였다. 센서 노드의 위치는 RSSI에 의해 선택된 가상 고정 노드에 WCL을 적용하여 결정한다. WCL에서 부여하는 가중치는 고정 노드와 RSSI 센서 노드 위치의 거리에 대한 함수이며, 고정 노드 간의 측정된 RSSI 값으로 구한다. 고정 노드 간의 RSSI로 가상 고정 노드를 이용한 WCL은 기존의 방식에서 사용되는 고정 노드의 수를 줄이고, 측정 대상의 위치 정확도를 측정된 RSSI 값만으로 추정한 위치보다 7.8% 향상되는 것을 확인 하였다.