• ETRI Journal
• /
• v.46 no.4
• /
• pp.671-682
• /
• 2024

Violence can be committed anywhere, even in crowded places. It is hence necessary to monitor human activities for public safety. Surveillance cameras can monitor surrounding activities but require human assistance to continuously monitor every incident. Automatic violence detection is needed for early warning and fast response. However, such automation is still challenging because of low video resolution and blind spots. This paper uses ResNet50v2 and the gated recurrent unit (GRU) algorithm to detect violence in the Movies, Hockey, and Crowd video datasets. Spatial features were extracted from each frame sequence of the video using a pretrained model from ResNet50V2, which was then classified using the optimal trained model on the GRU architecture. The experimental results were then compared with wavelet feature extraction methods and classification models, such as the convolutional neural network and long short-term memory. The results show that the proposed combination of ResNet50V2 and GRU is robust and delivers the best performance in terms of accuracy, recall, precision, and F1-score. The use of ResNet50V2 for feature extraction can improve model performance.

• Proceedings of the KSRS Conference
• /
• v.1
• /
• pp.29-33
• /
• 2006

The potential impact of past Caribbean tsunamis generated by earthquakes and/or massive submarine slides/slumps, as well as the tsunamigenic potential and population distribution within the Intra-Americas Sea (IAS) was examined to help define the optimal location for coastal sea level gauges intended to serve as elements of a regional tsunami warning system. The goal of this study was to identify the minimum number of sea level gauge locations to aid in tsunami detection and provide the most warning time to the largest number of people. We identified 12 initial, prioritized locations for coastal sea level gauge installation. Our study area approximately encompasses $7^{\circ}N$, $59^{\circ}W$ to $36^{\circ}N$, $98^{\circ}$ W. The results of this systematic approach to assess priority locations for coastal sea level gauges will assist in developing a tsunami warning system (TWS) for the IAS by the National Oceanic and Atmospheric Administration (NOAA) and the Intergovernmental Oceanographic Commission's Regional Sub-Commission for the Caribbean and Adjacent Regions (IOCARIBE-GOOS).

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.1
• /
• pp.33-40
• /
• 2012

This paper proposes a method to implement biological early warning systems(BEWS). This system generates periodically data event using a monitoring daemon and it extracts the feature parameters from this data sets. The feature parameters are derived with 6 variables, x/y coordinates, distance, absolute distance, angle, and fractal dimension. Specially by using the fractal dimension theory, the proposed algorithm define the input features represent the organism characteristics in non-toxic or toxic environment. And to find a moderate algorithm for learning the extracted feature data, the system uses an extended learning algorithm(UChoo) popularly used in machine learning. And this algorithm includes a learning method with the extended data expression to overcome the BEWS environment which the feature sets added periodically by a monitoring daemon. In this algorithm, decision tree classifier define class distribution information using the weight parameter in the extended data expression. Experimental results show that the proposed BEWS is available for environmental toxicity detection.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.14 no.9
• /
• pp.1979-1986
• /
• 2010

Biological early warning system detects toxicity by looking at behavior of organisms in water. The system uses classifier for judgement about existence and amount of toxicity in water. Boosting algorithm is one of possible application method for improving performance in a classifier. Boosting repetitively change training example set by focusing on difficult examples in basic classifier. As a result, prediction performance is improved for the events which are difficult to classify, but the information contained in the events which can be easily classified are discarded. In this paper, an incremental learning method to overcome this shortcoming is proposed by using the extended data expression. In this algorithm, decision tree classifier define class distribution information using the weight parameter in the extended data expression by exploiting the necessary information not only from the well classified, but also from the weakly classified events. Experimental results show that the new algorithm outperforms the former Learn++ method without using the weight parameter.

• The Journal of Society for e-Business Studies
• /
• v.12 no.3
• /
• pp.75-93
• /
• 2007

This paper presents real-time online early fire warning systems developed for preserving cultural properties of Bulguksa which is a world heritage designated by UNESCO. The system is based on the ubiquitous sensor network employing 900MHz and 2.4GHz bands. In this paper, we analyze requirements that should be considered in building effective management systems of cultural heritages by using wireless sensor network. Finally, we introduce the architecture, sensor and network design, and software design of the fire warning systems which is an initial version of U-Bulguksa. The current version of systems has been operating in Bukguksa for a few months. U-Bukguksa project sponsored by National Information Society Agency is ultimately aimed at developing an integrated system of U-cultural heritage management and U-tourism. The former aims to conserve and manage intangible cultural properties by providing a variety of environmental information such as erosion, crack, and gradient as well as fire which are important causes of loss and damage in real-time and online. The latter refers to the intelligent tourism information and guidance systems allowing tourists to get the personalized content on cultural heritages and help guidance with mobile devices in Bulguksa.

• Women's Health Nursing
• /
• v.6 no.1
• /
• pp.109-116
• /
• 2000

In modern society, the amount of alcohol ingestion is increasing at a dangerous level, especially among women. One of reason for increased alcohol consumption is stress caused by social pressures. Alcohol is a kind of depressant of centric nervous system, so it can induce relaxation of body and decrease the stress. The evidence on the effects of alcohol on the fetus is somewhat hazy, whereas that of smoking is quite clear. The literature on the ingestion of alcohol strongly suggests that drinking during pregnancy is associated with teratogenic effect and low birth weight. Therefore, the adverse effect of alcohol ingestion during pregnancy must be informed to public. More remarkable warning sign about alcohol ingestion must be attached on the top of bottles. This can be an effective measure for public education. Also legal sanction or tax imposition for the production of liquor be required. First of all, drinking habit or drinking culture must be changed. In fact, the strongest motivation of drinking in adolescent is a peer pressure which is related to drinking habit or culture. Secondly, early detection and treatment must be required to prevent from fetal alcohol syndrome. Accordingly, drinking history of pregnant women must be assessed as early as possible and health professional should give a warning about the abstinence of alcohol to drinking women. Thirdly, to minimize the adverse effects for mother, withdrawal syndrome by alcohol ingestion must be treated. to correct the malformation by fetal alcohol syndrome(FAS) can be corrected. Sometimes surgical intervention may be required for this purpose.

• Journal of the Society of Disaster Information
• /
• v.10 no.1
• /
• pp.105-115
• /
• 2014

In this research, we developed video analytic algorithms to detect water-level automatically and a system for proactive alarming using intelligent CCTV cameras. We applied these algorithms and a system to test-beds and verified for practical use. We made camera-selection policies and operation plans to keep the detection accuracy high and to optimize the suitability for the ever-changing weather condition, while the environmental factors such as camera shaking and weather condition can affect to detection accuracy. The estimation result of algorithms showed 90% detection accuracy for all CCTV camera types. For water level detection, NIR camera performed great. NIR camera performed over 95% accuracy in day or night, suitable in natural weather condition such as shaking condition, fog, and low light, needs similar installment skills with common cameras, and spends only 15% high cost. As a result, we practically tested water level detection algorithms and operation system based on intelligent CCTV camera. Furthermore, we expect the positive evidences when it is applied for public use.

• Smart Structures and Systems
• /
• v.30 no.6
• /
• pp.613-626
• /
• 2022

Guaranteeing the quality and integrity of structural health monitoring (SHM) data is very important for an effective assessment of structural condition. However, sensory system may malfunction due to sensor fault or harsh operational environment, resulting in multiple types of data anomaly existing in the measured data. Efficiently and automatically identifying anomalies from the vast amounts of measured data is significant for assessing the structural conditions and early warning for structural failure in SHM. The major challenges of current automated data anomaly detection methods are the imbalance of dataset categories. In terms of the feature of actual anomalous data, this paper proposes a data anomaly detection method based on data-level and deep learning technique for SHM of civil engineering structures. The proposed method consists of a data balancing phase to prepare a comprehensive training dataset based on data-level technique, and an anomaly detection phase based on a sophisticatedly designed network. The advanced densely connected convolutional network (DenseNet) and Transformer encoder are embedded in the specific network to facilitate extraction of both detail and global features of response data, and to establish the mapping between the highest level of abstractive features and data anomaly class. Numerical studies on a steel frame model are conducted to evaluate the performance and noise immunity of using the proposed network for data anomaly detection. The applicability of the proposed method for data anomaly classification is validated with the measured data of a practical supertall structure. The proposed method presents a remarkable performance on data anomaly detection, which reaches a 95.7% overall accuracy with practical engineering structural monitoring data, which demonstrates the effectiveness of data balancing and the robust classification capability of the proposed network.

• Proceedings of the Korea Society of Information Technology Applications Conference
• /
• 2005.11a
• /
• pp.123-126
• /
• 2005

Wireless Sensor Networks will revolutionize applications such as environmental monitoring, home automation, and logistics. We developed forest fire surveillance system. In this paper, Considering the fact that in Korea, during November to May, forest fires occur very frequently causing catastrophic damages on the valuable environment, Although exists other forest fire surveillance system such as surveillance camera tower, infrared ray sensor system and satellite system. Preexistence surveillance system can't real-time surveillance, monitoring, database and automatic alarm. But, forest fire surveillance system(FFSS) support above. In this paper, we describes a system development approach for a wireless sensor network based FFSS that is to be used to measure temperature and humidity as well as being fitted with a smoke detector. Such a device can be used as an early warning fire detection system and real-time surveillance in the area of a bush fire or endangered public infrastructure. Once the system has being development, a mesh network topology will be implemented with the chosen sensor node with the aim of developing a sophisticated mesh network.

• Journal of Korean Society for Quality Management
• /
• v.47 no.1
• /
• pp.75-86
• /
• 2019

Purpose: It is essential for the steel industry to produce steel products without unexpected downtime to reduce costs and produce high quality products. A hot strip rolling mill consists of many mechanical and electrical units. In condition monitoring and diagnosis, various units could fail for unknown reasons. Methods: In this study, we propose an effective method to detect units with abnormal status early to minimize system downtime. The early warning problem with various units was first defined. An autoencoder was modeled to detect abnormal states. An application of the proposed method was also implemented in a simulated field-data analysis. Results: We can compare images of original data and reconstructed images, as well as visually identify differences between original and reconstruction images. We confirmed that normal and abnormal states can be distinguished by reconstruction error of autoencoder. Experimental results show the possibility of prediction due to the increase of reconstruction error from just before equipment failure. Conclusion: In this paper, hot strip roughing mill monitoring method using autoencoder is proposed and experiments are performed to study the benefit of the autoencoder.