• 한국철도학회:학술대회논문집
• /
• 한국철도학회 2011년도 춘계학술대회 논문집
• /
• pp.1722-1728
• /
• 2011

The security issue in urban transit system has been widely considered as the common matters after the fire accident at Daegu subway station. The safe urban transit system is highly demanded because of the vast number of daily passengers, and it is one of the most challenging projects. We introduced a test model for integrated security system for urban transit system and built it at a subway station to demonstrate its performance. This system consists of cameras, sensor network and central monitoring software. We described the smart camera functionality in more detail. The proposed smart camera includes the moving objects recognition module, video analytics, video encoder and server module that transmits video and audio information.

• 대한의용생체공학회:학술대회논문집
• /
• 대한의용생체공학회 1996년도 추계학술대회
• /
• pp.48-51
• /
• 1996

We developed a multimedia monitoring system for patients at home, equipped with video and audio-conferencing capabilities. It can also monitor biological signals in real-time with vital signs. The system has an extensible architecture to accommodate probable future needs easily.

• 한국정보처리학회논문지
• /
• 제6권11S호
• /
• pp.3243-3250
• /
• 1999

There are many problems to be satisfied the various commercial demand in the present VRML node, and the VRML-based system by the contact with other technology is burdened with the direct participation of developer related with it. In this study, remote control and monitoring system using EAI(External Authoring Interface) is reposed to overcome the limit of VRML. In other words, This study is about mechanism of interaction between control and monitoring subject in real world and object in virtual world mapped with it. Object oriented method by one to one mapping of the Java class which have VRML virtual world and the object control is proposed. Introduction method of video monitoring through method of image synthesis and the integrated development environment is described. Finally automobile-gate control system based on the development method of Internet virtual reality-based system which is proposed in this study is described.

• 방송공학회논문지
• /
• 제25권1호
• /
• pp.13-24
• /
• 2020

한국에는 전력 분배를 위하여 약 9백만 개의 전신주와 1.3백만 킬로미터의 송전선이 있다. 이러한 많은 전력 설비의 유지보수를 위해서는 많은 인력과 시간이 소요된다. 최근 인공지능을 사용한 여러 고장진단 기술들이 연구되어 오고 있기 때문에 본 논문에서는 송전선의 여러 요인으로 인한 늘어짐을 감지하기 위해 기존의 현장에서의 검증 방법이 아닌 카메라 시스템으로 촬영한 영상에서의 인공 지능 기술을 활용한 송전선 늘어짐 감지 시스템을 제안한다. 제안하는 시스템은 (i) 객체 탐지 시스템을 이용한 송전탑 감지 (ii) 동영상 촬영 데이터의 화질 저하 문제를 해결하기 위한 히스토그램 평활화 기법 (iii) 송전선 전체를 파악하기 위한 파노라마 영상 스티칭(iv) 송전선 탐지 알고리즘 적용 후 파노라마 영상 스티칭 기술을 이용한 늘어짐 판단 과정으로 진행된다. 본 논문에서는 각각의 과정들에 대한 설명 및 실험 결과를 보인다.

• 디지털콘텐츠학회 논문지
• /
• 제17권4호
• /
• pp.273-278
• /
• 2016

재난은 다양성, 복잡성, 불가측성 등으로 현대사회의 특성과 유기적 관계가 있기 때문에 그 관리의 효율성을 위해 다양한 접근과 복합적인 처방으로 대국민에게 재난의 불안을 해소해 주어야 한다. 이에 따라 본 논문에서는 사회안전을 위한 지능형 영상감시 분석시스템의 구축방안과 이의 응용과 그 활용성의 장단점을 검토하였으며, 향후 제안 서비스가 사회안전을 위한 영상감시 시스템으로써 종합적인 도시 관제기능을 수행하면서 국민의 안전을 보장하고, 범죄와 사고를 예방하며, 범법행위를 사전에 단속하여 공공시설물과 국민의 재산을 보호할 것으로 기대된다.

• 한국생산제조학회지
• /
• 제26권3호
• /
• pp.272-277
• /
• 2017

CCTV image processing techniques have been developed for safety management in manufacturing sites. However, CCTV growth has become a social problem for video surveillance with regard to privacy. This study aims to manage the safety system efficiently and protect privacy simultaneously. In this study, the CCTV monitoring system is composed of five steps (accident monitoring, detection, notification, management, restoration). De-identified image is observed when we are in a normal situation. De-identified image changes to identified image when it detects an accident. As soon as it detects an accident, the accident information is sent to the safety administrator. Then the administrator could conduct safety measures. Afterward, accumulated accident data could be used for statistical data that could be utilized as analyzing expecting accident.

• 전기학회논문지P
• /
• 제62권4호
• /
• pp.164-168
• /
• 2013

Recently, in the elevator of the buildings, such as apartment and office building, the visitor monitoring cameras for surveillance, advertising LCD to provide a variety of information, emergency call devices, and safe driving information sensors are installed. A variety of multimedia data from these devices to the central control office, management office, or to guard room are transmitted in real-time. Each sub-systems in the elevator are installed with a separate lines and operated independently and use different management principals, so integrated management for each sub-systems are very difficult. In this study, we develop a new IP-based multi-channel video surveillance system which is integrated surveillance camera, emergency call devices, communications devices, various sensors in the elevator, DVR, ODM, and can manage all devices by two-way communication and integrated interface. And we evaluated the performance of the developed system.

• Journal of Ecology and Environment
• /
• 제42권2호
• /
• pp.85-89
• /
• 2018

Observations and video documentation of interactions between animals living in dens, cavities, and other enclosed spaces are difficult, but they play an important role in field biology, ecology, and conservation. For example, bird parents visiting nests and feeding their nestlings may provide crucial information for testing of ecological hypotheses and may easily attract attention of participants of citizen-science ecological and conservation projects. Because of the nest concealment of cavity-nesting birds, their behaviors in the nest can only be studied by using video surveillance. Professional wildlife surveillance systems are extremely expensive. Here, we describe an inexpensive video setup that can be constructed with relatively little effort and is more affordable than any previously described system. We anticipate that the relatively low cost of about 250 USD for a battery-operated system is an important feature for citizen-science type of projects and for applications in heavily populated areas where the potential for theft and vandalism may be high. Based on our experiences, we provide methodological advice on practical aspects of using this system in the field for ecological research on birds. We highlight the low cost, easiness of construction, and potential availability to a large number of observers taking part in wildlife monitoring projects, and we offer technical help to participants of such research projects.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2002년도 하계학술대회 논문집 C
• /
• pp.1649-1651
• /
• 2002

Ultra-wide bandwidth (UWB) detection method has been widely used as a front end detection scheme for the ultra-high frequency (UHF) partial discharge (PD) monitoring systems. A broad-band video detector module was developed and characterized for the UHF UWB PD detection systems. The useable bandwidth of the module is more than 2 GHz and it is optimized for 50-ohm systems. The detection sensitivity and dynamic range of the module were characterized by using a known ns-width RF pulses of GHz range. The dynamic range is more than 6-decades and the module can detect pulsed RF signals down to 1 nW. The detector module can eliminate expensive equipment such as high speed oscilloscopes and radio frequency (RF) spectrum analyzers. Therefore, it enables one to use slow speed data acquisition systems for the PD monitoring at the UHF range. The detector module was used to detect real PDs of about <3 pC. The module converts the UWB PD signals into a low-bandwidth video signal with a high signal-to-noise ratio.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2021년도 추계학술발표대회
• /
• pp.704-707
• /
• 2021

Pigs' aggressive behavior represents one of the common issues that occur inside pigpens and which harm pigs' health and welfare, resulting in a financial burden to farmers. Continuously monitoring several pigs for 24 hours to identify those behaviors manually is a very difficult task for pig caretakers. In this study, we propose a lightweight video-based approach for monitoring pigs' aggressive behavior that can be implemented even in small-scale farms. The proposed system receives sequences of frames extracted from an RGB video stream containing pigs and uses MnasNet with a DM value of 0.5 to extract image features from pigs' ROI identified by predefined annotations. These extracted features are then forwarded to a lightweight LSTM to learn temporal features and perform behavior recognition. The experimental results show that our proposed model achieved 0.92 in recall and F1-score with an execution time of 118.16 ms/sequence.