• Journal of KIISE:Software and Applications
• /
• v.31 no.9
• /
• pp.1161-1170
• /
• 2004

This paper proposes to segment skin color areas using a clustering algorithm. Most of previously proposed clustering algorithms have some difficulties, since they generally detect hyperspherical clusters, run in a batch mode, and predefine a number of clusters. In this paper, we use a well-known elliptical clustering algorithm, an EM algorithm, and modify it to learn on-line and find automatically the number of clusters, called to an EAM algorithm. The effectiveness of the EAM algorithm is demonstrated on a task of skin color region segmentation. Experimental results present the EAM algorithm automatically finds a right number of clusters in a given image without any information on the number. Comparing with the EM algorithm, we achieved better segmentation results with the EAM algorithm. Successful results were achieved to detect and segment skin color regions using a conditional probability on a region. Also, we applied to classify images with persons and got good classification results.

• The Korean Journal of Applied Statistics
• /
• v.33 no.2
• /
• pp.115-122
• /
• 2020

K-means clustering uses a spherical or elliptical metric to group data points; however, it does not work well for non-convex data such as the concentric circles. Spectral clustering, based on graph theory, is a generalized and robust technique to deal with non-standard type of data such as non-convex data. Results obtained by spectral clustering often outperform traditional clustering such as K-means. In this paper, we review spectral clustering and show important issues in spectral clustering such as determining the number of clusters K, estimation of scale parameter in the adjacency of two points, and the dimension reduction technique in clustering high-dimensional data.

• Journal of Institute of Control, Robotics and Systems
• /
• v.9 no.7
• /
• pp.536-546
• /
• 2003

We propose an object extraction technique adequate for the radial shape's radar signal structure for the purpose of implementing ARPA(Automatic Radar Plotting Aid) installed in the vessel. The radar signal data are processed by interpolation and accumulation to acquire a qualified image. The objects of the radar image have characteristics of having different shape and size as it gets far from the center, and it is not adequate for clustering generally. Therefore, this study designs a new vigilance distance model of elliptical shape and adopts this model in the ART2 neural network. We prove that the proposed clustering method makes it possible to extract objects adaptively and to separate the connected objects effectively.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.12C
• /
• pp.1262-1267
• /
• 2005

Elliptical basis function (EBF) neural network works stably under high-level background noise environment and makes the nonlinear processing possible. It can be adapted real time VAD with simple design. This paper introduces VAD implementation using EBF and the experimental results show that EBF VAD outperforms G729 Annex B and RBF neural networks. The best error rates achieved by the EBF networks were improved more than $70\%$ in speech and $50\%$ in silence while that achieved by G.729 Annex B and RBF networks respectively.

• Journal of The Korean Astronomical Society
• /
• v.50 no.3
• /
• pp.61-70
• /
• 2017

We study the angular correlation function of bright ($K_s{\leq}19.5$) Extremely Red Objects (EROs) selected in the Subaru GTO 2$deg^2$ field. By applying the color selection criteria of $R-K_s$ > 5.0, 5.5, and 6.0, we identify 9055, 4270, and 1777 EROs, respectively. The number density is consistent with similar studies on the optical - NIR color selected red galaxies. The angular correlation functions are derived for EROs with different limiting magnitude and different $R-K_s$ color cut. When we assume that the angular correlation function $w({\theta})$ follows a form of a power-law (i.e., $w({\theta})=A{\theta}^{-{\delta}}$), the value of the amplitude A was larger for brighter EROs compared to the fainter EROs. The result suggests that the brighter, thus more massive high-redshift galaxies, are clustered more strongly compared to the less massive galaxies. Assuming that EROs have redshift distribution centered at ~ 1.1 with ${\sigma}_z=0.15$, the spatial correlation length $r_0$ of the EROs estimated from the observed angular correlation function ranges ${\sim}6-10h^{-1}Mpc$. A comparison with the clustering of dark matter halos in numerical simulation suggests that the EROs are located in most massive dark matter halos and could be progenitors of $L_{\ast}$ elliptical galaxies.

• The Bulletin of The Korean Astronomical Society
• /
• v.36 no.2
• /
• pp.51.2-51.2
• /
• 2011

We have studied the topology of volume-limited galaxy sample selected from the very luminous red galaxies (LRGs) in the completed Sloan Digital Sky Survey. LRGs are predominantly massive elliptical galaxies and tend to reside in massive dark matter halos. We compared the observed genus statistics with predictions from perturbation theory and mock LRG surveys constructed from dark matter halos in a Lambda CDM model. To compare with the observational data, we made 129 mock surveys in the past light cone space by using three different size CDM simulations: 41203 particle 6592 Mpc/h, 60003 particle 7200 Mpc/h, and $7210^3$ particle 10815 Mpc/h.

• The Korean Journal of Mycology
• /
• v.23 no.1 s.72
• /
• pp.92-104
• /
• 1995

Ten species of Cordyceps species were collected throughout Kangwon province including Chuncheon Dongsanmyun KNU forest experiment from June to September, 1993. Collected Cordyceps species were identified as Cordyceps militaris, C. roseostromata, C. kyushuensis, C. scarabaeicola, Phytocordyceps ninchukiospora, C. nutans, Paecilomyces tenuipes, C. sphecocephala, Hymenostilbe odonatae, Torrubiella sp.. C. militaris, type species of Cordyceps species, was mainly formed on pupae of Lepidoptera and found after the rainy season around July. Fruiting body of C. roseostromata was morphologically similar to those of C. militaris, but relatively small in size and they were also found on lawn or pupa of Lepidoptera. Fruiting body of C. scarabaeicola was found on adult Scarabaeidae specifically and collect fruiting bodies of C. kyushuensis were on larva of moth. C. nutans and C. sphecocephala had host specificity on Hemiptera and Hymenoptera, respectively. Each species formed elliptical fertile part attach to the slim and carneous stalk and they were collected the most in specimen number through whole season of the summer. Ascospore of Phytocordyceps ninchukiospora on seed was characterized by two viable, multiseptate, fusiform units linked end-to-end by a long, filiform connective. Paecilomyces tenuipes, imperfect stage of the genus Cordyceps is multi-infective fungi that attack all stages of all groups of insects. Hymenostilbe odonatae attacks only adult Odonata and Torrubiella sp. formed on spider was difficult to collect because it was found the back side of leaf. As results of cultural test PDA medium showed the best mycelial growth. In the experiment of effect of the acidity inside of the media, C. militaris was good on pH 5, C. nutans and Phytocordyceps ninchukiospora were good on pH 6 and Paecilomyces tenuipes was on pH 7 and C. scarabaeicola was on pH 9. All isolates tested showed the best mycelial growth at $20^{\circ}C$. Morphologically similar isolates were used to analyze protein banding pattern among and within species. As a result, C. militaris, C. roseostromata and C. kyushuensis were clustered into close species and C. scarabaeicola and Phytocordyceps ninchukiospora were relatively distant from those species.