• Proceedings of the KSRS Conference
• /
• 1998.09a
• /
• pp.201-206
• /
• 1998

In this paper, we have considered the method for clustering land cover types over the East Asia from AVHRR data. The feature vectors such that maximum NDVI, amplitude of NDVI, mean NDVI, and NDVI threshold are extracted from the 10-day composite by maximum value composite(MVC) for reducing the effect of cloud contaninations. To find the land cover clusters given by the feature vectors, we are adapted the self-organizing feature map(SOFM) clustering which is the mapping of an input vector space of n-dimensions into a one - or two-dimensional grid of output layer. The approach is to find first the clusters by the first layer SOFM and then merge several clusters of the first layer to a large cluster by the second layer SOFM. In experiments, we were used the 8-km AVHRR data for two years(1992-1993) over the East Asia.

• Journal of Mechanical Science and Technology
• /
• v.19 no.8
• /
• pp.1682-1691
• /
• 2005

Direct numerical simulation database of an axial turbulent boundary layer is used to compute frequency and wave number spectra of the wall shear-stress fluctuations in a low-Reynolds number axial turbulent boundary layer. One-dimensional and two-dimensional power spectra of flow variables are calculated and compared. At low wave numbers and frequencies, the power of streamwise shear stress is larger than that of spanwise shear stress, while the powers of both stresses are almost the same at high wave numbers and frequencies. The frequency/streamwise wave number spectra of the wall flow variables show that large-scale fluctuations to the rms value is largest for the stream wise shear stress, while that of small-scale fluctuations to the rms value is largest for pressure. In the two-point auto-correlations, negative correlation occurs in streamwise separations for pressure, and in span wise correlation for both shear stresses.

• Proceedings of the KIEE Conference
• /
• 1994.07b
• /
• pp.1478-1480
• /
• 1994

In this paper, the microwave absorption properties of two layered absorber were evaluated. The two layer absorber were composed of Ni-Zn ferrite/rubber absorber and Ferroxplana/rubber absorber. The most absorption properties condition were obtained when the transforming layer and absorbing layer were Ferroxplana/robber absorber and Ni -Zn ferrite/rubber absorber. respectivity. The optimal thickness of the transforming and the absorbing layer were 4(mm) and 1(mm), respective1y.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.3B
• /
• pp.565-569
• /
• 2000

We compose the two-stage recurrent neural network that returns both signals of a hidden and an output layer to the hidden layer. It is tested on the basis of syllables for Korean spoken digit from /gong/to /gu. For these experiments, we adjust the neuron number of the hidden layer, the predictive order of input data and self-recurrent coefficient of the decision state layer. By the experimental results, the recognition rate of this neural network is between 91% and 97.5% in the speaker-dependent case and between 80.75% and 92% in the speaker-independent case. In the speaker-dependent case, this network shows an equivalent recognition performance to Jordan and Elman network but in the speaker-independent case, it does improved performance.

• Proceedings of the Korean Society of Tribologists and Lubrication Engineers Conference
• /
• 2001.06a
• /
• pp.401-405
• /
• 2001

The effects of oxide layer formed on TiN coated ball and counter-body have been investigated from the frictional point of view during sliding tests. AISI52100 steel ball was used for the substrate of coated specimens. Two types of coated specimens were prepared by depositing TiN coating with 1 and 4$\mu\textrm{m}$ in coating thickness. AISI1045 steel was used for the disk type counter-body. To investigate the effect of oxide layer on the contact parts of two materials, the tests were performed both in ambient for forming oxide layer on the contact parts and in nitrogen environment to avoid oxidation.

• The Journal of the Korean Society for Microbiology
• /
• v.8 no.1
• /
• pp.33-36
• /
• 1973

This study was carried out for the classification of mycobacteria using the thin-layer chromatography of mycobacterial lipids. Results as follows: Of the 12 strains of mycobacteria, the two spots on chromatogarphy were three strains of mycobacteria(BCG, M. tuberculosis($H_{37}J$) and M. ulcerans) and three spots on chromatography were two strains of mycobacteria(M. kansasii and M. balnei). The method of thin-layer chromatography of mycobacterial lipids was considered to capable for the use of classification of mycobacteria.

• Korean Journal of Air-Conditioning and Refrigeration Engineering
• /
• v.15 no.11
• /
• pp.895-901
• /
• 2003

Direct numerical simulation database of an axial turbulent boundary layer is used to compute frequency and wave number spectra of the wall shear-stress fluctuations in a low-Reynolds number axial turbulent boundary layer. One-dimensional and two-dimensional power spectra of flow variables are calculated and compared. At low wave numbers and frequencies, the power of streamwise shear stress is larger than that of spanwise shear stress, while the powers of both stresses are almost the same at high wave numbers and frequencies. The frequency/streamwise wave number spectra of the wall flow variables show that large-scale fluctuations to the ms value is largest for the streamwise shear stress, while that of small-scale fluctuations to the rms value is largest for pressure. In the two-point auto-correlations, negative correlation occurs in streamwise separations for pressure and spanwise shear stress, and in spanwise correlation for both shear stresses.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.23 no.4
• /
• pp.550-556
• /
• 1986

This paper proposes a two-layer channel routing algorithm using the division of signal nets in LSI/VLSI layout design. To solve the vertical constraint problem, the doglegging method is used. Although signal net division and the dogleg are used, the routing is accomplished within local channel density and the increase in vias is repressed by assining the vertical segments to the metal layer and the horizontal segments to the poly layer. The algorithm was implemented on a VAX 11/780 computer. The effectiveness of the proposed algorithm is proved by appling this algorithm to Deutch's difficult example.

• The Journal of the Acoustical Society of Korea
• /
• v.25 no.2E
• /
• pp.74-78
• /
• 2006

This paper proposes an effective method for classifying emotions of the music from its acoustical signals. Two feature sets, timbre and tempo, are directly extracted from the modified discrete cosine transform coefficients (MDCT), which are the output of partial MP3 (MPEG 1 Layer 3) decoder. Our tempo feature extraction method is based on the long-term modulation spectrum analysis. In order to effectively combine these two feature sets with different time resolution in an integrated system, a classifier with two layers based on AdaBoost algorithm is used. In the first layer the MDCT-driven timbre features are employed. By adding the MDCT-driven tempo feature in the second layer, the classification precision is improved dramatically.

• Journal of computational fluids engineering
• /
• v.2 no.1
• /
• pp.129-137
• /
• 1997

We consider long nonlinear waves in the two-layer flow of an inviscid and incompressible fluid bounded above by a free surface and below by a rigid boundary. The flow is forced by a bump on the bottom. The derivation of the forced KdV equation fails when the density ratio h and the depth ratio ρ yields a condition 1＋hρ=(2-h)((1-h)²＋4ρh)/sup 1/2/. To overcome this difficulty we derive a forced modified KdV equation by a refined asymptotic method. Numerical solutions are given and hydraulic fall solution of a two layer fluid is expressed analytically in the case that derivation of the forced KdV(FKdV) equaition fails.