• Journal of the Korea Society of Computer and Information
• /
• v.18 no.6
• /
• pp.21-28
• /
• 2013

This paper proposes an effective fire detection approach that includes the following multiple heterogeneous algorithms: moving region detection using grey level histograms, color segmentation using fuzzy c-means clustering (FCM), feature extraction using a grey level co-occurrence matrix (GLCM), and fire classification using support vector machine (SVM). The proposed approach determines the optimal threshold values based on grey level histograms in order to detect moving regions, and then performs color segmentation in the CIE LAB color space by applying the FCM. These steps help to specify candidate regions of fire. We then extract features of fire using the GLCM and these features are used as inputs of SVM to classify fire or non-fire. We evaluate the proposed approach by comparing it with two state-of-the-art fire detection algorithms in terms of the fire detection rate (or percentages of true positive, PTP) and the false fire detection rate (or percentages of true negative, PTN). Experimental results indicated that the proposed approach outperformed conventional fire detection algorithms by yielding 97.94% for PTP and 4.63% for PTN, respectively.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.13 no.3
• /
• pp.197-205
• /
• 2020

Korean language has the characteristics that the pronunciation of phoneme units such as vowels and consonants are fixed and the pronunciation associated with a notation does not change, so that foreign learners can approach rather easily Korean language. However, when one pronounces words, phrases, or sentences, the pronunciation changes in a manner of a wide variation and complexity at the boundaries of syllables, and the association of notation and pronunciation does not hold any more. Consequently, it is very difficult for foreign learners to study Korean standard pronunciations. Despite these difficulties, it is believed that systematic analysis of pronunciation errors for Korean words is possible according to the advantageous observations that the relationship between Korean notations and pronunciations can be described as a set of firm rules without exceptions unlike other languages including English. In this paper, we propose a visualization framework which shows the differences between standard pronunciations and erratic ones as quantitative measures on the computer screen. Previous researches only show color representation and 3D graphics of speech properties, or an animated view of changing shapes of lips and mouth cavity. Moreover, the features used in the analysis are only point data such as the average of a speech range. In this study, we propose a method which can directly use the time-series data instead of using summary or distorted data. This was realized by using the deep learning-based technique which combines Self-organizing map, variational autoencoder model, and Markov model, and we achieved a superior performance enhancement compared to the method using the point-based data.

• The KIPS Transactions:PartD
• /
• v.16D no.3
• /
• pp.307-316
• /
• 2009

In this paper, the spatiotemporal data mining methodology for detecting a cycle of power consumption pattern with the change of time and spatial was proposed, and applied to the power consumption data collected by GIS-AMR system with an aim to use its resulting knowledge in real world applications. First, partial clustering method was applied for cluster analysis concerned with the aim of customer's power consumption. Second, the patterns of customer's power consumption data which contain time and spatial attribute were detected by 3D cube mining method. Third, using the calendar pattern mining method for detection of cyclic patterns in the various time domains, the meanings and relationships of time attribute which is previously detected patterns were analyzed and predicted. For the evaluation of the proposed spatiotemporal data mining, we analyzed and predicted the power consumption patterns included the cycle of time and spatial feature from total 266,426 data of 3,256 customers with high power consumption from Jan. 2007 to Apr. 2007 supported by the GIS-AMR system in KEPRI. As a result of applying the proposed analysis methodology, cyclic patterns of each representative profiles of a group is identified on time and location.

• The KIPS Transactions:PartB
• /
• v.10B no.1
• /
• pp.67-72
• /
• 2003

In this paper, we study the enhancement of VQ (Vector Quantization) design for text dependent speaker recognition. In a concrete way, we present the non-Iterative method which makes a vector quantization codebook and this method Is nut Iterative learning so that the computational complexity is epochally reduced. The proposed semi-noniterative VQ design method contrasts with the existing design method which uses the iterative learning algorithm for every training speaker. The characteristics of a semi-noniterative VQ design is as follows. First, the proposed method performs the iterative learning only for the reference speaker, but the existing method performs the iterative learning for every speaker. Second, the quantization region of the non-reference speaker is equivalent for a quantization region of the reference speaker. And the quantization point of the non-reference speaker is the optimal point for the statistical distribution of the non-reference speaker In the numerical experiment, we use the 12th met-cepstrum feature vectors of 20 speakers and compare it with the existing method, changing the codebook size from 2 to 32. The recognition rate of the proposed method is 100% for suitable codebook size and adequate training data. It is equal to the recognition rate of the existing method. Therefore the proposed semi-noniterative VQ design method is, reducing computational complexity and maintaining the recognition rate, new alternative proposal.

• Journal of Life Science
• /
• v.18 no.4
• /
• pp.565-571
• /
• 2008

Due to rapid development of bioinformatics technologies, various biological data have been produced in silico. So now days complicated and large scale biodata are used to accomplish requirement of researcher. Developing visualization and annotation tool using them is still hot issues although those have been studied for a decade. However, diversity and various requirements of users make us hard to develop general purpose tool. In this paper, I propose a novel system, Genome Viewer and Annotation tool (GenoVA), to annotate and visualize among genomes using known information and multiple relation graph. There are several multiple alignment tools but they lose conserved area for complexity of its constrains. The GenoVA extracts all associated information between all pair genomes by extending pairwise alignment. High frequency conserved area and high BLAST score make a block node of relation graph. To represent multiple relation graph, the system connects among associated block nodes. Also the system shows the known information, COG, gene and hierarchical path of block node. In this case, the system can annotates missed area and unknown gene by navigating the special block node's clustering. I experimented ten bacteria genomes for extracting the feature to visualize and annotate among them. GenoVA also supports simple and rough computational annotation of new genome.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.55 no.3
• /
• pp.217-232
• /
• 2019

It was turned out by shrimp beam trawl monthly survey from March, 2015 to February, 2016 that different species composition and abundance of the fish assemblages in Sacheon Bay and coastal waters off Namhae, Korea were compared. As a result of monthly measured sea temperature and salinity of Sacheon Bay and coastal waters off Namhae, sea temperature of both areas was changed seasonally; however, differences in sea temperature occurred during certain periods depending on the region. Salinity was generally low in Sacheon Bay affected by fresh water, and both areas was low in summer and high in winter. A total of 73 species representing 37 families were collected in Sacheon Bay. The dominant fish species in terms of numbers and biomass were Liparis tanakae, $23,077inds./km^2$, $332.1kg/km^2$. A total of 91 fish species representing 49 families were collected in coastal waters off Namhae. The dominant fish species in terms of numbers were Leiognathus nuchalis, $139,683inds./km^2$ and biomass were Chelidonichthys spinosus, $1,078.6kg/km^2$. Analysis of dendrogram of the clustering showed that Sacheon Bay and coastal waters off Namhae were distinctive featured (global R = 0.691, p = 0.017). And except of summer season (July-October), there was a distinctive feature seasonally (global R = 0.844, p = 0.001). The fish species that appeared in common in both areas, where fish species caught in Sacheon Bay, an important inner bay,were smaller than those caught in coastal waters off Namhae appeared. It presented that Sacheon Bay plays a more important role in spawning and nursery ground for fisheries resource than coastal waters off Namhae, Korea.

• Journal of KIISE:Computer Systems and Theory
• /
• v.33 no.3
• /
• pp.157-165
• /
• 2006

The change of external/internal factors of the cell rquires specific biological functions to maintain life. Such functions encourage particular genes to jnteract/regulate each other in multiple ways. Accordingly, we applied a linear decomposition model IFSA, which derives hidden variables, called the 'expression mode' that corresponds to the functions. To interpret gene interaction/regulation, we used a cross-correlation method given an expression mode. Linear decomposition models such as principal component analysis (PCA) and independent component analysis (ICA) were shown to be useful in analyzing high dimensional DNA microarray data, compared to clustering methods. These methods assume that gene expression is controlled by a linear combination of uncorrelated/indepdendent latent variables. However these methods have some difficulty in grouping similar patterns which are slightly time-delayed or asymmetric since only exactly matched Patterns are considered. In order to overcome this, we employ the (IFSA) method of [1] to locate phase- and shut-invariant features. Membership scoring functions play an important role to classify genes since linear decomposition models basically aim at data reduction not but at grouping data. We address a new function essential to the IFSA method. In this paper we stress that IFSA is useful in grouping functionally-related genes in the presence of time-shift and expression phase variance. Ultimately, we propose a new approach to investigate the multiple interaction information of genes.