• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.11a
• /
• pp.224-227
• /
• 2020

엣지컴퓨팅 기반 유해조수 퇴치 Drone의 유해조수 추적 기술은 Doppler Sensor를 이용해 사유지에 침입한 유해조수를 인식 후 사용자에게 위험 요소에 대한 알림 서비스를 제공한다. 이후 사용자는 Drone의 Camera와 전용 애플리케이션을 이용해 경작지를 실시간으로 보며 Drone을 조종한다. Camera는 Tensor Flow Object Detection Deep Learning을 적용하여 유해조수를 학습 및 파악, 추적한다. 이후 Drone은 Speaker와 Neo Pixel LED Ring을 이용해 유해조수의 시각과 청각을 자극해 도망을 유도하며 퇴치한다. Tensor Flow object detection을 핵심으로 Drone에 접목했고 이를 위해 전용 애플리케이션을 개발했다.

• International Journal of Computer Science & Network Security
• /
• v.23 no.6
• /
• pp.115-120
• /
• 2023

Plant leaf diseases and destructive insects are major challenges that affect the agriculture production of the country. Accurate and fast prediction of leaf diseases in crops could help to build-up a suitable treatment technique while considerably reducing the economic and crop losses. In this paper, Convolutional Neural Network based model is proposed to detect leaf diseases of a plant in an efficient manner. Convolutional Neural Network (CNN) is the key technique in Deep learning mainly used for object identification. This model includes an image classifier which is built using machine learning concepts. Tensor Flow runs in the backend and Python programming is used in this model. Previous methods are based on various image processing techniques which are implemented in MATLAB. These methods lack the flexibility of providing good level of accuracy. The proposed system can effectively identify different types of diseases with its ability to deal with complex scenarios from a plant's area. Predictor model is used to precise the disease and showcase the accurate problem which helps in enhancing the noble employment of the farmers. Experimental results indicate that an accuracy of around 93% can be achieved using this model on a prepared Data Set.

• International Journal of Computer Science & Network Security
• /
• v.24 no.4
• /
• pp.107-112
• /
• 2024

Plant leaf diseases and destructive insects are major challenges that affect the agriculture production of the country. Accurate and fast prediction of leaf diseases in crops could help to build-up a suitable treatment technique while considerably reducing the economic and crop losses. In this paper, Convolutional Neural Network based model is proposed to detect leaf diseases of a plant in an efficient manner. Convolutional Neural Network (CNN) is the key technique in Deep learning mainly used for object identification. This model includes an image classifier which is built using machine learning concepts. Tensor Flow runs in the backend and Python programming is used in this model. Previous methods are based on various image processing techniques which are implemented in MATLAB. These methods lack the flexibility of providing good level of accuracy. The proposed system can effectively identify different types of diseases with its ability to deal with complex scenarios from a plant's area. Predictor model is used to precise the disease and showcase the accurate problem which helps in enhancing the noble employment of the farmers. Experimental results indicate that an accuracy of around 93% can be achieved using this model on a prepared Data Set.

• Korea-Australia Rheology Journal
• /
• v.18 no.3
• /
• pp.161-167
• /
• 2006

This study presents results on the numerical simulation of Newtonian and non-Newtonian flow in a channel obstructed by an asymmetric array of obstacles for clarifying the descriptive ability of current non-Newtonian constitutive equations. Jones and Walters (1989) have performed the corresponding experiment that clearly demonstrates the characteristic difference among the flow patterns of the various liquids. In order to appropriately account for flow properties, the Navier-Stokes, the Carreau viscous and the Leonov equations are employed for Newtonian, shear thinning and extension hardening liquids, respectively. Making use of the tensor-logarithmic formulation of the Leonov model in the computational scheme, we have obtained stable solutions up to relatively high Deborah numbers. The peculiar characteristics of the non-Newtonian liquids such as shear thinning and extension hardening seem to be properly illustrated by the flow modeling. In our opinion, the results show the possibility of current constitutive modeling to appropriately describe non-Newtonian flow phenomena at least qualitatively, even though the model parameters specified for the current computation do not precisely represent material characteristics.

• Geophysics and Geophysical Exploration
• /
• v.17 no.2
• /
• pp.58-65
• /
• 2014

We studied the source mechanisms of very-long-period event recorded at seismic station WIZ near White Island Volcano, New Zealand on August 4, 2012. Since seismic data at only one station were available, we conducted moment tensor inversion using three simplified models (explosion, crack, and pipe models). To constrain the moment tensor solution of seismic event, we computed synthetic data for each model to compare with observed data. Type and orientation for the best model is a crack at a depth of 1600 m with a dip of $80^{\circ}NE$ and a strike of $N80^{\circ}W$. We interpret that a deep explosion may have opened a crack for gases to escape, and the upward gas flow triggered the surface explosions four hours later as confirmed by a webcam. The interpretation based on moment tensor inversion is consistent with previous studies of geochemical data of the volcanic island.

• Bulletin of the Korean Mathematical Society
• /
• v.48 no.6
• /
• pp.1315-1327
• /
• 2011

Let M be a real hypersurface of a complex space form with almost contact metric structure (${\phi}$, ${\xi}$, ${\eta}$, g). In this paper, we study real hypersurfaces in a complex space form whose structure Jacobi operator $R_{\xi}=R({\cdot},\;{\xi}){\xi}$ is ${\xi}$-parallel. In particular, we prove that the condition ${\nabla}_{\xi}R_{\xi}=0$ characterizes the homogeneous real hypersurfaces of type A in a complex projective space or a complex hyperbolic space when $R_{\xi}{\phi}S=R_{\xi}S{\phi}$ holds on M, where S denotes the Ricci tensor of type (1,1) on M.

• Honam Mathematical Journal
• /
• v.36 no.4
• /
• pp.805-812
• /
• 2014

For unit tangent sphere bundles $T_1M$ with the standard contact metric structure (${\eta},\bar{g},{\phi},{\xi}$), we have two fundamental operators that is, $h=\frac{1}{2}{\pounds}_{\xi}{\phi}$ and ${\ell}=\bar{R}({\cdot},{\xi}){\xi}$, where ${\pounds}_{\xi}$ denotes Lie differentiation for the Reeb vector field ${\xi}$ and $\bar{R}$ denotes the Riemmannian curvature tensor of $T_1M$. In this paper, we study the Reeb ow invariancy of the corresponding (0, 2)-tensor fields H and L of h and ${\ell}$, respectively.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.19 no.3
• /
• pp.820-833
• /
• 1995

An analysis is performed to examine the equilibrium state and the stability of modeled Reynolds stress equations for homogeneous turbulent shear flows. The system of the governing equations consists of four coupled ordinary differential equations. The equilibrium states are found by the steady state solution of the governing equations. In order to investigate the stability of the system about its state in equilibrium, and eigenvalue problem is constructed. As a result, constraints for the coeffieients in the model equations are obtained by the stability condition of the equilibrium state as well as by their physically realizable bounds. It is observed that the models with pressure-strain rate correlation that are linear in the anisotropy tensor are stable and produce reasonable equilibrium tensor do not behave properly. Stability considerations about three most commonly used models are given in detail in the final section.

• Journal of computational fluids engineering
• /
• v.2 no.2
• /
• pp.9-17
• /
• 1997

The performance of several turbulence models in computing an axisymmetric supersonic base flow is investigated. A compressible Navier-Stokes code, which incorporates k-ε, k-ω model and Reynolds stress closure with three kinds of pressure-strain correlation model, has been developed using implicit LU-SGS algorithm with second-order upwind TVD scheme. Numerical computations have been carried out for Herrin and Dutton's base flow. It is observed that the two-equation models give large backward axial velocity approaching to the base and somewhat larger variation of base pressure distribution than the Reynolds stress model. It is also found that the Reynolds stress model with third order pressure-strain model in the anisotropy tensor predicts most accurate mean flow field.

• Korea-Australia Rheology Journal
• /
• v.19 no.4
• /
• pp.211-219
• /
• 2007

Here we demonstrate complex transient behavior of viscoelastic liquid described numerically with the Leonov model in straight and contraction channel flow domains. Finite element and implicit Euler time integration methods are employed for spatial discretization and time marching. In order to stabilize the computational procedure, the tensor-logarithmic formulation of the constitutive equation with SUPG and DEVSS algorithms is implemented. For completeness of numerical formulation, the so called traction boundaries are assigned for flow inlet and outlet boundaries. At the inlet, finite traction force in the flow direction with stress free condition is allocated whereas the traction free boundary is assigned at the outlet. The numerical result has illustrated severe forward-backward fluctuations of overall flow rate in inertial straight channel flow ultimately followed by steady state of forward flow. When the flow reversal occurs, the flow patterns exhibit quite complicated time variation of streamlines. In the inertialess flow, it takes much more time to reach the steady state in the contraction flow than in the straight pipe flow. Even in the inertialess case during startup contraction flow, quite distinctly altering flow patterns with the lapse of time have been observed such as appearing and vanishing of lip vortices, coexistence of multiple vortices at the contraction comer and their merging into one.