• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.21 no.1
• /
• pp.151-156
• /
• 2017

In case of massive water leakage, it's crucial for field manager to quickly positioning the problematic valve and related ones. However, it's not easy for the system to find the corresponding valve and even if it's found, it can not respond quickly because it can't know the relevant information immediately. In this paper, we implement the system for identifying sluice valve positions using GPS and AR techniques. The proposed system is composed of hand held android device, remote database server and data acquisition device for DB creation. We utilize the android device's sensors including GPS, gyro, accelerometer, magnetic sensor. The system identifies the valve with matching between the position data from the remote database server, and current GPS locations of device. We use AR techniques to overlay the graphics pattern of valve positions and some additional informations on captured real scene. With this system, it will be fast and accurate for maintenance of sluice valve of municipal water system.

• Journal of the Korean Society of Hazard Mitigation
• /
• v.9 no.6
• /
• pp.11-15
• /
• 2009

Recently, a new PSC (Prestressed Concrete) bridge system, which is supported by Concrete-filled fiber-reinforced polymer (CFFRP) strut, has been introduced. This bridge is able to reduce self-weight and increase the width of traditional PSC bridges. However, no relevant research has been reported on local behavior of CFFRP strut in the bridge system. The purpose of this study is to investigate local behavior of CFFRP struts using fiber Bragg grating (FBG) sensors. Field tests were performed to examine the hoop strains and longitudinal strains of the FRP strut under various lateral positions and velocities of a test truck. It has been observed that CFFRP strut is under compression regardless of vehicle speed and location. However, the CFFRP strut is sensitive to the lateral position of vehicles in terms of strain magnitude. Results also indicated that the FBG sensors can faithfully record the hoop and longitudinal strains of the FRP strut without electro-magnetic interference.

• Journal of Sensor Science and Technology
• /
• v.6 no.4
• /
• pp.258-264
• /
• 1997

$YBa_{2}Cu_{3}O_{7}$ single layer dc SQUID magnetometers, prepared on $1\;cm^{2}\;SrTiO_{3}$ substrates, have been fabricated and characterized. Based on the analytical description, a SQUID magnetometer design having a 8.5 mm pickup coil with 2.6 mm linewidth, and a SQUID inductance Ls = 50 pH with $3\;{\mu}m$ Josephson junctions is presented. The devices showed a maximum modulation voltage depth of $65\;{\mu}V$ and a magnetic field noise of 0.6 pT /$\sqrt{Hz}$ at 1 Hz. Clear traces of human magnetocardiogram could be obtained with the SQUID magnetometer operating at 77 K.

• Korean Journal of Remote Sensing
• /
• v.33 no.5_3
• /
• pp.855-866
• /
• 2017

Several research cases using remote sensing methods to analyze changes of storage and dynamics of groundwater aquifer were reviewed in this paper. The status of groundwater storage, in an area with regional scale, could be qualitatively inferred from geological feature, surface water altimetry and topography, distribution of vegetation, and difference between precipitation and evapotranspiration. These qualitative indicators could be measured by geological lineament analysis, airborne magnetic survey, DEM analysis, LAI and NDVI calculation, and surface energy balance modeling. It is certain that GRACE and InSAR have received remarkable attentions as direct utilization from satellite data for quantification of groundwater storage and dynamics. GRACE, composed of twin satellites having acceleration sensors, could detect global or regional microgravity changes and transform them into mass changes of water on surface and inside of the Earth. Numerous studies in terms of groundwater storage using GRACE sensor data were performed with several merits such that (1) there is no requirement of sensor data, (2) auxiliary data for quantification of groundwater can be entirely obtained from another satellite sensors, and (3) algorithms for processing measured data have continuously progressed from designated data management center. The limitations of GRACE for groundwater storage measurement could be defined as follows: (1) In an area with small scale, mass change quantification of groundwater might be inaccurate due to detection limit of the acceleration sensor, and (2) the results would be overestimated in case of combination between sensor and field survey data. InSAR can quantify the dynamic characteristics of aquifer by measuring vertical micro displacement, using linear proportional relation between groundwater head and vertical surface movement. However, InSAR data might now constrain their application to arid or semi-arid area whose land cover appear to be simple, and are hard to apply to the area with the anticipation of loss of coherence with surface. Development of GRACE and InSAR sensor data preprocessing algorithms optimized to topography, geology, and natural conditions of Korea should be prioritized to regionally quantify the mass change and dynamics of the groundwater resources of Korea.

• Journal of Astronomy and Space Sciences
• /
• v.25 no.4
• /
• pp.435-444
• /
• 2008

High energy charged particles are trapped by geomagnetic field in the region named Van Allen Belt. These particles can move to low altitude along magnetic field and threaten even low altitude spacecraft. Space Radiation can cause equipment failures and on occasions can even destroy operations of satellites in orbit. Sun sensors aboard Science and Technology Satellite (STSAT-l) was designed to detect sun light with silicon solar cells which performance was degraded during satellite operation. In this study, we try to identify which particle contribute to the solar cell degradation with ground based radiation facilities. We measured the short circuit current after bombarding electrons and protons on the solar cells same as STSAT-1 sun sensors. Also we estimated particle flux on the STSAT-l orbit with analyzing NOAA POES particle data. Our result clearly shows STSAT-l solar cell degradation was caused by energetic protons which energy is about 700keV to 1.5MeV. Our result can be applied to estimate solar cell conditions of other satellites.

• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.163-177
• /
• 2019

As smartphones are getting widely used, human activity recognition (HAR) tasks for recognizing personal activities of smartphone users with multimodal data have been actively studied recently. The research area is expanding from the recognition of the simple body movement of an individual user to the recognition of low-level behavior and high-level behavior. However, HAR tasks for recognizing interaction behavior with other people, such as whether the user is accompanying or communicating with someone else, have gotten less attention so far. And previous research for recognizing interaction behavior has usually depended on audio, Bluetooth, and Wi-Fi sensors, which are vulnerable to privacy issues and require much time to collect enough data. Whereas physical sensors including accelerometer, magnetic field and gyroscope sensors are less vulnerable to privacy issues and can collect a large amount of data within a short time. In this paper, a method for detecting accompanying status based on deep learning model by only using multimodal physical sensor data, such as an accelerometer, magnetic field and gyroscope, was proposed. The accompanying status was defined as a redefinition of a part of the user interaction behavior, including whether the user is accompanying with an acquaintance at a close distance and the user is actively communicating with the acquaintance. A framework based on convolutional neural networks (CNN) and long short-term memory (LSTM) recurrent networks for classifying accompanying and conversation was proposed. First, a data preprocessing method which consists of time synchronization of multimodal data from different physical sensors, data normalization and sequence data generation was introduced. We applied the nearest interpolation to synchronize the time of collected data from different sensors. Normalization was performed for each x, y, z axis value of the sensor data, and the sequence data was generated according to the sliding window method. Then, the sequence data became the input for CNN, where feature maps representing local dependencies of the original sequence are extracted. The CNN consisted of 3 convolutional layers and did not have a pooling layer to maintain the temporal information of the sequence data. Next, LSTM recurrent networks received the feature maps, learned long-term dependencies from them and extracted features. The LSTM recurrent networks consisted of two layers, each with 128 cells. Finally, the extracted features were used for classification by softmax classifier. The loss function of the model was cross entropy function and the weights of the model were randomly initialized on a normal distribution with an average of 0 and a standard deviation of 0.1. The model was trained using adaptive moment estimation (ADAM) optimization algorithm and the mini batch size was set to 128. We applied dropout to input values of the LSTM recurrent networks to prevent overfitting. The initial learning rate was set to 0.001, and it decreased exponentially by 0.99 at the end of each epoch training. An Android smartphone application was developed and released to collect data. We collected smartphone data for a total of 18 subjects. Using the data, the model classified accompanying and conversation by 98.74% and 98.83% accuracy each. Both the F1 score and accuracy of the model were higher than the F1 score and accuracy of the majority vote classifier, support vector machine, and deep recurrent neural network. In the future research, we will focus on more rigorous multimodal sensor data synchronization methods that minimize the time stamp differences. In addition, we will further study transfer learning method that enables transfer of trained models tailored to the training data to the evaluation data that follows a different distribution. It is expected that a model capable of exhibiting robust recognition performance against changes in data that is not considered in the model learning stage will be obtained.

• Progress in Superconductivity
• /
• v.7 no.1
• /
• pp.28-35
• /
• 2005

We propose a method to measure atrial arrhythmias (AA) such as atrial fibrillation (Afb) and atrial flutter (Afl) with a SQUID magnetocardiograph (MCG) system. To detect AA is one of challenging topics in MCG. As the AA generally have irregular rhythm and atrio-ventricular conduction, the MCG signal cannot be improved by QRS averaging; therefore a SQUID MCG system having a high SNR is required to measure informative atrial excitation with a single scan. In the case of Afb, diminished f waves are much smaller than normal P waves because the sources are usually located on the posterior wall of the heart. In this study, we utilize an MCG system measuring tangential field components, which is known to be more sensitive to a deeper current source. The average noise spectral density of the whole system in a magnetic shielded room was $10\;fT/{\surd}Hz(a)\;1\;Hz\;and\;5\;fT/{\surd}Hz\;(a)\;100\;Hz$. We measured the MCG signals of patients with chronic Afb and Afl. Before the AA measurement, the comparison between the measurements in supine and prone positions for P waves has been conducted and the experiment gave a result that the supine position is more suitable to measure the atrial excitation. Therefore, the AA was measured in subject's supine position. Clinical potential of AA measurement in MCG is to find an aspect of a reentry circuit and to localize the abnormal stimulation noninvasively. To give useful information about the abnormal excitation, we have developed a method, separative synthetic aperture magnetometry (sSAM). The basic idea of sSAM is to visualize current source distribution corresponding to the atrial excitation, which are separated from the ventricular excitation and the Gaussian sensor noises. By using sSAM, we localized the source of an Afl successfully.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2017.05a
• /
• pp.255-257
• /
• 2017

These days automotive markets are launching models that include a variety of IT technologies. Tesla's Tesla model S and Google's unmanned automobiles are emerging one after another. This type of automobile with IT technology provides various convenience to the driver and the driver is getting benefit by various conveience services. on the contrary, it is also true that defects for errors in electronic components cause accidents that threaten the safety of drivers. There is a sudden unintended acceleration among these accidents. The cause of the accident is not clear yet, but the claim that the ECU device caused by the magnetic field causes accident of the car due is the most reliable. But, in Korea, when occur a car sudden unintended acceleration accident, the char maker often claims that an accident occurred due to driver's pedal malfunction. Also most drivers are responsible for the lack of grounds to refute. In this paper, the pedal operation image of the driver is acquired and the sensor is attached to the control part such as the excel and brake so as to discriminate whether the vehicle sudden unintended acceleration accident is the driver's pedal operation error or the fault of. i have implemented a system that can do this.