• Journal of the Korea Society of Computer and Information
• /
• v.17 no.10
• /
• pp.35-41
• /
• 2012

We propose a new marker driven Watershed algorithm for automated segmentation of clustered cell from microscopy image with less over segmentation. The Watershed Transform is able to segment extremely complex objects which are highly touched and overlapped each other. The success of the Watershed Transform depends essentially on the finding markers for each of the objects of interest. For extracting of markers positioning around center of each cell we used radial symmetry and iterative voting algorithms. With synthetic and real images, we quantitatively demonstrate the performance of our method and achieved better results than the other compared methods.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.38 no.6
• /
• pp.640-647
• /
• 2001

In this paper, a noise-tolerant generalized symmetry transform(NTGST) is proposed as an effective attention operator for the spot detection in BLU inspection, in which various spots with variable sizes, shapes, gray levels, and low contrast, should be detected from the complex, noisy background with lattice shaped shading. The proposed NTGST takes into account the polarity of convergence and divergence of the radial orientation of the intensity gradient as well as it's magnitude and symmetry, and thereby can detect only the BLU spots from the noisy and lattice shaped shadows of background. Experiments are conducted on the BLU inspection image obtained by CCD camera, and the proposed NTGST is Proved to be effectively used in BLU inspection.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.64 no.7
• /
• pp.1104-1113
• /
• 2015

In this paper, we propose Two-Dimensional Robust Face Recognition System Realized with the Aid of Facial Symmetry with Illumination Variation. Preprocessing process is carried out to obtain mirror image which means new image rearranged by using difference between light and shade of right and left face based on a vertical axis of original face image. After image preprocessing, high dimensional image data is transformed to low-dimensional feature data through 2-directional and 2-dimensional Principal Component Analysis (2D)2PCA, which is one of dimensional reduction techniques. Polynomial-based Radial Basis Function Neural Network pattern classifier is used for face recognition. While FCM clustering is applied in the hidden layer, connection weights are defined as a linear polynomial function. In addition, the coefficients of linear function are learned through Weighted Least Square Estimation(WLSE). The Structural as well as parametric factors of the proposed classifier are optimized by using Particle Swarm Optimization(PSO). In the experiment, Yale B data is employed in order to confirm the advantage of the proposed methodology designed in the diverse illumination variation

• Proceedings of the Korean Society for Technology of Plasticity Conference
• /
• 2005.10a
• /
• pp.95-98
• /
• 2005

Shafts having asymmetry or odd number of symmetry in the cross-section can not be simply manufactured by conventional incremental radial forging. In order to manufacture such shafts, the new concept of incremental forging with one punch and a flexible fixture is developed by suggesting a flexible fixture, instead of two opposed punches used in radial forging, so that the flexible fixture only supports the workpiece while the punch is moving during forming. A new flexible fixture is designed using the steel shots and vacuum technology. An equilateral triangular cross-section is selected as the sample shape to be manufactured by the proposed manufacturing method. The desired triangular cross-sectional shaft is manufactured with the errors of $3.0\%$.

• Bulletin of the Korean Chemical Society
• /
• v.7 no.4
• /
• pp.299-304
• /
• 1986

Master formulas for overlap integrals and the dipole moments involving |4p > atomic orbitals have been derived by the expansion method for spherical harmonics. The radial integrals for the hyperfine interaction have also been derived for |4p > atomic orbitals. The calculated values of the overlap integrals and dipole moment matrix elements by the expansion method for spherical harmonics for a hypothetical NO molecule are exactly in agreement with those of Mulliken's method. The radial integrals for the hyperfine interaction may be used to calculate the chemical shift for |4p > atomic orbitals.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.15 no.5
• /
• pp.841-848
• /
• 2020

The 1.29 GHz wind profiler radar is a remote observation device that is useful not only for calculating wind vectors in clear air, but also for detecting rainfall. The Doppler spectrum symmetry test is essential in the horizontal wind treatment process. Since asymmetry may be detected in rainfall cases, it is necessary to reflect in the wind calculation algorithm that the sign of the radial velocity is the same according to the magnitude of the vertical velocity. In the summer of 2017 (June, July), a wind vector calculation algorithm by Bragg scattering and Rayleigh scattering was developed using Changwon wind profiler data, and verified by comparing it with radiosonde data at 6 hour intervals.

• Bulletin of the Korean Chemical Society
• /
• v.5 no.2
• /
• pp.55-60
• /
• 1984

The NMR shift arising from the electron angular momentum and electron spin dipolar-nuclear spin angular momentum interactions has been investigated for a $4d^1$system in a strong crystal field environment of tetragonal symmetry. A general formula for NMR shift is used to compute the NMR shifts along the (100), (010), (001), (110) and (111) axes. We find that from the computed results, the NMR shift along the (100) and (010) axes is consistent with each other in a strong crystal field environment of tetragonal symmetry, but the NMR shift along the (001) axis is about triply greater in magnitude than those along the (100) and (010) axes and is opposite in sign to those along (100) and (010) axes. In this work, we express the expansion coefficients $a_1^{(i)}$ and $b_1^{(i)}$ of $A_i$ and $B_i$ in terms of $g_m^{(i)}$ and $h_m^{(i)}$ and two matrices $c_{lm}$ and $d_{lm}$ of radial dependence. The NMR shift is also separated into the contributions of multipolar terms. We find that $1/R^3$ term contributes dominantly to the NMR shift along the (100), (010), (001) and (110) axes while along the (111) axis $1/R^5$ term dominantly contributes. However, the contribtions of the other terms may not be negligible.

• Proceedings of the KSR Conference
• /
• 2005.05a
• /
• pp.459-466
• /
• 2005

In this paper, we have systematically formulated the equations concerned to the in-plane and out-of-plane motions and deformations of a thick circular beam by using the kinetic and strain energies in order to analyse natural frequencies of a thick ring. The effects of variation of radius of curvature across the cross-section and also the effects of bending shear, extension and twist are considered. The equations of motion for natural vibration analysis of a ring are obtained utilizing the cyclic symmetry of vibration modes of the ring. The frequencies calculated using thick ring model and thin ring model are compared and discussed with the ones obtained from finite element analysis using the method of cyclic symmetry with 20-node hexahedral solid elements for rings with the different ratio of radial thickness to mean radius.

• Korean Journal of Remote Sensing
• /
• v.24 no.5
• /
• pp.473-481
• /
• 2008

This paper presents an approach to Back-propagation and Radial Basis Function neural network method with various training set for automatic vehicle detection from aerial images. The initial extraction of candidate object is based on Mean-shift algorithm with symmetric property of a vehicle structure. By fusing the density and the symmetry, the method can remove the ambiguous objects and reduce the cost of processing in the next stage. To extract features from the detected object, we describe the object as a log-polar shape histogram using edge strengths of object and represent the orientation and distance from its center. The spatial histogram is used for calculating the momentum of object and compensating the direction of object. BPNN and RBFNN are applied to verify the object as a vehicle using a variety of non-car training sets. The proposed algorithm shows the results which are according to the training data. By comparing the training sets, advantages and disadvantages of them have been discussed.

• Journal of ILASS-Korea
• /
• v.5 no.4
• /
• pp.40-46
• /
• 2000

To understand the basic the structure of the spray field and to obtain the initial conditions for computational models for shear coaxial twin-fluid injectors. the atomization characteristics under different flow and geometric conditions were examined. The spray characteristics such as SMD, mean axial and radial velocities, Dia. of droplets and volume flux with a P.D.P.A. Water and nitrogen gas under atmospheric conditions were used as a test fluids. The drops produced by shear coaxial injectors continue to disintegrate along the spray axis and decrease their sizes. SMD was the maximum at the spray center of spray and decreased with increasing radial distance. The results of this parametric study showed that SMD decreased with increasing gas injection velocity as well as with decreasing liquid injection mass flow rate, The relative velocity between gas and liquid flow played a significant role resulted in decreasing SMD and in spreading the spray. Recessing the liquid orifice resulted decreasing SMD and a spreading the spray. Recess of liquid orifice by 5.0mm showed best atomization characteristics in this experiment. Although drop diameter changes, shear coaxial injector sprays had constant velocity and exhibited a high degree of radial symmetry.