• Journal of the Society of Disaster Information
• /
• v.18 no.2
• /
• pp.364-373
• /
• 2022

Purpose: The purpose of this paper is to develope smoke detection using AI model for detecting the initial fire in underground utility tunnels using CCTV Method: To improve detection performance of smoke which is high irregular, a deep learning model for fire detection was trained to optimize smoke detection. Also, several approaches such as dataset cleansing and gradient exploding release were applied to enhance model, and compared with results of those. Result: Results show the proposed approaches can improve the model performance, and the final model has good prediction capability according to several indexes such as mAP. However, the final model has low false negative but high false positive capacities. Conclusion: The present model can apply to smoke detection in underground utility tunnel, fixing the defect by linking between the model and the utility tunnel control system.

• Proceedings of the KSME Conference
• /
• 2008.11a
• /
• pp.684-689
• /
• 2008

In this study, homogenization method combined with the effective interface model for the characterization of properties of the nanoparticulate composites is developed. In order to characterize particle size effect of nanocomposites, effective interface model has been developed. The application range of analytical micromechanics approach is limited because a simple analytical approach is valid only for simple and uniform geometry of fiber particles. Therefore this study focuses on the analysis of mechanical properties of the effect interface through the continuum homogenization method instead of using analytical micromechanics approach. Using the homogenization method, elastic stiffness properties of the effective interface are numerically evaluated and compared with the analytically obtained micromechanics solutions. The suggested homogenization method is expected to be applied to optimization problems for nanocomposite design.

• Journal of Biomedical Engineering Research
• /
• v.5 no.2
• /
• pp.161-166
• /
• 1984

In this paper, we describe an effective technique for computing the steady-state motion in a two-dimensional cochlear model. With the cochlear fluid assumed incompressible and invisid, the problem reduces to solving an integral equation for a region with yielding boundary. Using the conformal mapping, Jacobian elliptic function and hyperbolic function, a pair of second-order differential equation is derived. What we will show in this paper is that by appropriately transforming integral equation, the same computation can be performed with comparable accuracy in a short time.

• Journal of Biomedical Engineering Research
• /
• v.9 no.1
• /
• pp.1-10
• /
• 1988

In this study, the recognition experiments for Korean vowel are performed using peripheral auditory model. In addition, for the purpose of objective comparison, the recognition experiments are performed by extracting LPC cepstrum coefficients for the same speech data. The results are as follows. 1) The time and the frequency responses of the auditory model show that important features of input signal are involved in the responses of inner ear and auditory nerve. 2) The recognition results for Korean vowel show that the recognition rate by auditory model output is higher than the recognition rate by LPC cepstrum coefficients. 3) The adaptation phenomenon of auditory nerve provides useful characteristics for the discrimination of vowel signal.

• Proceedings of the KIEE Conference
• /
• 1998.07g
• /
• pp.2486-2488
• /
• 1998

In this paper, Prameters esimated t using AR model in order to approach linearly the heart sound which include the nonlinear characteristic from the characteristics based on a statistical theory. The parameters which is figured out using AR model is a very important information which show the characteristic heart sound In this paper parameters estimated using autocorrelation method and order selected by proposed Akaike[6] method. Compared the similirities of the spectrums between estimated by using AR model and estimated by using FFT method.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.20 no.3
• /
• pp.1096-1105
• /
• 1996

A theoretical model of the frost formation has been presented to investigate the characteristics of the growth of frost layer by considering molecular diffusion of water vapor and heat generation caused by sublimation of water vapor in the frost layer. The present model was compared with existing experimental data as well as a previous model. The difference between the present model and existing experimental data was found to be about 6 percent. An analysis for the behavior of frost formation using present model shows a good agreement with the trend for a number of experimentally observed features. The present analysis can also provide the physical understanding on the phenomena of the frost formation.

• Composites Research
• /
• v.30 no.4
• /
• pp.247-253
• /
• 2017

In this study, the mechanical behavior and interface properties of boron nitride nanotube-poly(methyl methacrylate) nanocomposites are predicted using the molecular dynamics simulations and the double inclusion model. After modeling nanocomposite unit cell embedding single-walled nanotube and polymer, the stiffness matrix is determined from uniaxial tension and shear tests. Through the orientation average of the transversely isotropic stiffness matrix, the effective isotropic elastic constants of randomly dispersed microstructure of nanocomposites. Compared with the double inclusion model solution with a perfect interfacial condition, it is found that the interface between boron nitride nanotube and polymer matrix is weak in nature. To characterize the interphase surrounding the nanotube, the two step domain decomposition method incorporating a linear spring model at the interface is adopted. As a result, various combinations of the interfacial compliance and the interphase elastic constants are successfully determined from an inverse analysis.

• Composites Research
• /
• v.31 no.6
• /
• pp.371-378
• /
• 2018

In this study, the thermoelastic properties of poly lactic acid (PLA) based nanocomposites are predicted by molecular dynamics (MD) simulation and a micromechanics model. The stereocomplex mixed with L-lactic acid (PLLA) and D-lactic acid (PDLA) is modeled as matrix phase and a single walled carbon nanotube is embedded as reinforcement. The glass transition temperature, elastic moduli and thermal expansion coefficients of pure matrix and nanocomposites unit cells are predicted though ensemble simulations according to the hydrolysis. In micromechanics model, the double inclusion (D-I) model with a perfect interface condition is adopted to predict the properties of nanocomposites at the same composition. It is found that the stereocomplex nanocomposites show prominent improvement in thermal stability and interfacial adsorption regardless of the hydrolysis. Moreover, it is confirmed from the comparison of MD simulation results with those from the D-I model that the interface between CNT and the stereocomplex matrix is slightly weak in nature.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.12
• /
• pp.320-330
• /
• 2020

Fire is difficult to achieve good performance in image detection using deep learning because of its high irregularity. In particular, there is little data on fire detection in underground utility facilities, which have poor light conditions and many objects similar to fire. These make fire detection challenging and cause low performance of deep learning models. Therefore, this study proposed a fire detection model using deep learning and estimated the performance of the model. The proposed model was designed using a combination of a basic convolutional neural network, Inception block of GoogleNet, and Skip connection of ResNet to optimize the deep learning model for fire detection under underground utility facilities. In addition, a training technique for the model was proposed. To examine the effectiveness of the method, the trained model was applied to fire images, which included fire and non-fire (which can be misunderstood as a fire) objects under the underground facilities or similar conditions, and results were analyzed. Metrics, such as precision and recall from deep learning models of other studies, were compared with those of the proposed model to estimate the model performance qualitatively. The results showed that the proposed model has high precision and recall for fire detection under low light intensity and both low erroneous and missing detection capabilities for things similar to fire.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.54 no.8
• /
• pp.123-132
• /
• 2017

Traffic lights contain very important information for safety driving. So, the delivery of the information to drivers in real-time is a very critical issue for advanced driver assistance systems. However, traffic light detection is quite difficult because of the small sized traffic lights and the occlusion in real world. In this paper, a traffic light detection method using modified color based saliency map and morphological information is proposed. It shows 98.14% of precisions and 83.52% of recalls on computer simulations.