• 한국멀티미디어학회논문지
• /
• 제24권2호
• /
• pp.186-198
• /
• 2021

Security control using surveillance cameras is established when people observe all surveillance videos directly. However, this task is labor-intensive and it is difficult to detect all abnormal situations. In this paper, we propose a deep neural network model, called AT-Net, that automatically detects abnormal situations in the surveillance video, and introduces an automatic video surveillance system developed based on this network model. In particular, AT-Net alleviates the ambiguity of existing abnormal situation detection methods by mapping features representing relationships between people and objects in surveillance video to the new tensor structure based on sparse coding. Through experiments on actual surveillance videos, AT-Net achieved an F1-score of about 89%, and improved abnormal situation detection performance by more than 25% compared to existing methods.

• 한국컴퓨터정보학회논문지
• /
• 제27권4호
• /
• pp.111-118
• /
• 2022

최근 1인 고령 가구가 증가하고 있지만 1인 가구의 경우 집 안에서 위험 상황이 발생했을 때, 이를 외부에 알리기 힘들다. 이와 같은 1인 가구의 위험 상황을 탐지하기 위해 다양한 스마트홈 솔루션이 제안되고 있지만, 프라이버시 영역에 문제가 있는 홈 CCTV와 같은 영상 매체는 활용 어렵다. 그리고 단일 센서만을 활용하여 집안 내 고령자의 위험 상황을 분석할 경우, 데이터양의 한계로 정확한 상황해석이 제한 된다. 따라서 본 논문에서는 프라이버시를 지킬 수 있으며 실생활에 밀접한 2DLiDAR, 먼지, 음성 센서 간의 상관관계 따른 융합을 통한 집 내부의 위험 상황 탐지 융합 알고리즘을 제안한다. 또한, 본 논문은 실제 환경에서 수집한 데이터를 통해 알고리즘의 신뢰성을 증명한다. 제안하는 알고리즘이 탐지 가능한 위험 상황과 불가능한 상황을 제시한다. 본 논문은 집 안에서 위험 상황을 탐지하는 연구로써 1인 가구 사용자의 생활에 도움이 될 것이다.

• 전자공학회논문지SC
• /
• 제48권5호
• /
• pp.25-30
• /
• 2011

본 논문은 교통 감시를 수행하는 고정 카메라에서, 움직이는 물체들의 궤적을 사용자가 입력한 사용자 지정 경로를 바탕으로 그 정상/비정상성을 판별하는 방법을 제안한다. 제안된 방법은 입력된 경로 정보를 미리 정해진 규칙에 따라 각각의 이동 물체에 대한 비정상성(abnormality)을 계산하고 이를 임계값(Threshold)과 비교하여 비정상 행위를 판별해낸다. 사용자의 경로 정보 입력 기능을 이용하기 때문에 기존의 방법들에서 사용한, 계산량과 시간 소모가 크며 학습 데이터에 의해 그 성능이 크게 영향을 받는 정상 행위 (normal behavior) 모델링 단계를 배제하여 보다 빠르고 정확한 판별 결과를 제공한다. 뿐만 아니라 단순히 지정된 규칙만을 이용하지 않고 주어진 환경에 따라 규칙을 변형 적용하여 보다 강인한 판별 결과를 제공한다. 실험 결과는 본 논문에서 제안한 방법이 각종 교통 상황에서 발생하는 불법 및 비정상 교통 행위를 강인하게 판별해 냄을 보여준다.

• 한국멀티미디어학회논문지
• /
• 제20권2호
• /
• pp.216-224
• /
• 2017

Abnormal situation caused by aggressive behavior of pigs adversely affects the growth of pigs, and comes with an economic loss in intensive pigsties. Therefore, IT-based video surveillance system is needed to monitor the abnormal situations in pigsty continuously in order to minimize the economic demage. Recently, some advances have been made in pig monitoring; however, detecting each pig is still challenging problem. In this paper, we propose a new color image-based monitoring system for the detection of the individual pig using a fast region-based convolution neural network with consideration of detecting touching pigs in a crowed pigsty. The experimental results with the color images obtained from a pig farm located in Sejong city illustrate the efficiency of the proposed method.

• 대한의용생체공학회:의공학회지
• /
• 제25권2호
• /
• pp.129-135
• /
• 2004

재택건강관리 시스템은 주로 정상인들로부터 빈번하게 측정한 생체신호의 실시간 처리과정을 필요로 한다. 본 논문에서는 이러한 환경에서 측정되는 심전도 신호에서 DRS를 검출하기 위한 단순화된 알고리즘과 검출된 QRS의 정상과 비정상 여부만을 분류하는 알고리즘에 대하여 기술한다. 기존에 사용되고 있는 실시간 QRS 검출 알고리즘을 세분화하여 단순화된 QRS 검출 알고리즘을 제안함으로서 저가형 소형 단말기에서도 사용이 가능하도록 하였다. 또한 검출된 QRS들로부터 QRS 폭, R-R 간격, DRS 형태변수를 추출하여 QRS의 정상과 비정상을 판단하는 알고리즘을 개발하였다. 단순화된 QRS 검출기의 성능과 정상과 비정상의 분류성능은 각각 약 99%와 96%로 나타났다. 본 논문에서 제안된 QRS 검출과 분류를 위한 알고리즘들은 복잡한 신호처리 과정이 필요치 않으므로 재택건강관리 시스템에서의 실시간 심전도처리에 사용될 수 있을 것이다

• 산업공학
• /
• 제20권3호
• /
• pp.288-297
• /
• 2007

Success of semiconductor/LCD industry depends on its yield and quality of product. For the purpose, FDC (Fault Detection and Classification) system is used to diagnose fault state in main manufacturing processes by monitoring time series data collected by equipment sensors which represent various conditions of the equipment. The data set is segmented at the start and end of each product lot processing by a trigger event module. However, in practice, segmented sensor data usually have the features of data asynchronization such as different start points, end points, and data lengths. Due to the asynchronization problem, false alarm (type I error) and missed alarm (type II error) occur frequently. In this paper, we propose a robust process fault detection system by integrating a process event detection method and a similarity measuring method based on dynamic time warping algorithm. An experiment shows that the proposed system is able to recognize abnormal condition correctly under the asynchronous data situation.

• Journal of Positioning, Navigation, and Timing
• /
• 제11권2호
• /
• pp.109-117
• /
• 2022

This paper proposes a method for detecting flight anomalies of drones through the difference between the command of flight controller (FC) and the navigation solution. If the drones make a flight normally, control errors generated by the difference between the desired control command of FC and the navigation solution should converge to zero. However, there is a risk of sudden change or divergence of control errors when the FC control feedback loop preset for the normal flight encounters interferences such as strong winds or navigation sensor abnormalities. In this paper, we propose the method with a deep neural network model that predicts the control error in the normal flight so that the abnormal flight state can be detected. The performance of proposed method was evaluated using the real-world flight data. The results showed that the method effectively detects anomalies in various situation.

• 전기학회논문지
• /
• 제67권12호
• /
• pp.1619-1625
• /
• 2018

Portable electronic devices secondary batteries can cause fire and explosion due to micro-current change in addition to the situation of short-circuit inrush current, safety can not be secured with a general operation limited current fuse. Therefore, in secondary battery, it is necessary for the protector to satisfy both the limit current type operation in the open-short-circuit inrush current and the current detection operation characteristic in the micro current change situation and for this operation, a fuse for the current detection type secondary battery protection circuit can be applied. The purpose of this study is to design a protection device that operates stably in the hazardous situation of small capacity secondary battery for portable electronic devices through the design of low melting fuse elements alloy of sensing type fuse and secures stability in abnormal current state. As a result of the experiment, I-T and V-T operation characteristics are satisfied in a the design of the alloy of the current sensing type self-contained low melting point fuse and the resistance of the heating resistor. It is confirmed that it can prevent accidents of short circuit over-current and micro current change of secondary battery.

• International journal of advanced smart convergence
• /
• 제4권2호
• /
• pp.46-53
• /
• 2015

This paper suggests a method of real time confidence interval estimation to detect abnormal states of sensor data. For real time confidence interval estimation, the mean square errors of the exponential smoothing method and moving average method, two of the time series analysis method, were compared, and the moving average method with less errors was applied. When the sensor data passes the bounds of the confidence interval estimation, the administrator is notified through alarms. As the suggested method is for real time anomaly detection in a ship, an Android terminal was adopted for better communication between the wireless sensor network and users. For safe navigation, an administrator can make decisions promptly and accurately upon emergency situation in a ship by referring to the anomaly detection information through real time confidence interval estimation.

• 대한임베디드공학회논문지
• /
• 제19권1호
• /
• pp.57-64
• /
• 2024

In the industrial field, robots are used to increase productivity by replacing labors with dangerous, difficult, and hard tasks. However, failures of individual industrial robots in the entire production process may cause product defects or malfunctions, and may cause dangerous disasters in the case of manufacturing parts used in automobiles and aircrafts. Although requirements for early diagnosis of industrial robot failures are steadily increasing, there are many limitations in early detection. This paper introduces methods for diagnosing robot failures using sound-based data and deep learning. This paper also analyzes, compares, and evaluates the performance of failure diagnosis using various deep learning technologies. Furthermore, in order to improve the performance of the fault diagnosis system using deep learning technology, we propose a method to increase the accuracy of fault diagnosis based on an inference window. When adopting the inference window of deep learning, the accuracy of the failure diagnosis was increased up to 94%.