• Journal of the Korea Society of Computer and Information
• /
• v.19 no.12
• /
• pp.11-19
• /
• 2014

In this paper, we design an intelligent agent system with the data mining module and information fusion module as the core components of the system and investigate the possibility for the medical expert system. In the data mining module, fuzzy neural network, OFUN-NET analyzes multi-view data and produces fuzzy cluster knowledge base. In the information fusion module and application module, they serve the diagnosis result with possibility degree and useful information for diagnosis, such as uncertainty decision status or detection of asymmetry. We also present the experiment results on the BI-RADS-based feature data set selected form DDSM benchmark database. They show higher classification accuracy than conventional methods and the feasibility of the system as a computer aided diagnosis system.

• Nuclear Engineering and Technology
• /
• v.45 no.6
• /
• pp.795-802
• /
• 2013

For the improvement of APR1400 Diverse Protection System (DPS) design, the Advanced DPS (ADPS) has recently been developed to enhance the fault tolerance capability of the system. Major fault masking features of the ADPS compared with the APR1400 DPS are the changes to the channel configuration and reactor trip actuation equipment. To minimize the fault occurrences within the ADPS, and to mitigate the consequences of common-cause failures (CCF) within the safety I&C systems, several fault avoidance design features have been applied in the ADPS. The fault avoidance design features include the changes to the system software classification, communication methods, equipment platform, MMI equipment, etc. In addition, the fault detection, location, containment, and recovery processes have been incorporated in the ADPS design. Therefore, it is expected that the ADPS can provide an enhanced fault tolerance capability against the possible faults within the system and its input/output equipment, and the CCF of safety systems.

• Proceedings of the Korean Geotechical Society Conference
• /
• 2001.03a
• /
• pp.249-256
• /
• 2001

Optical Color Corescanner firstly developed by DMT-GeoTec, Germany and further upgraded through the Korea-Germany joint project is capable of duplicating the core surfaces. The tool uses a digital CCD line camera. As the core is rotated by an electric motor, the camera scans the uppermost line, everytime with a circumferential increment of up to 0.05mm(20pixels/mm) and hence a complete 360$^{\circ}$ unwrapped image(core image) is produced. This paper illustrated diverse research benefits of such core images from several test sites in our country. All scanned images could be stored as a data-base one and easily used with software facilities \circled1 to evaluate a percental distribution of mineral components or grain size etc. not only for the rock classification but also for e.g. the assessment of building stones, \circled2 to study potential reservoirs as a hydrocarbon indicator using ultraviolet fluorescence reflection from cores, \circled3 to facilitate the qualitative and quantitative analysis of fractures, \circled4 to evaluate the fractures and thin bedded reservoirs using spectral color responses. Based on abundant scanning experiments, it would seem that this imaging work should lead to reflecting the future trend in underground survey toward a more comprehensive understanding of the properties and behaviors of in situ rocks.

• Journal of Advanced Marine Engineering and Technology
• /
• v.39 no.9
• /
• pp.903-910
• /
• 2015

The purpose of this research is to develop a full-spectral fatigue analysis program, based on rigid-body ship motion analysis, in order to perform a full-spectral fatigue analysis that considers hydroelasticity effects. To gain credibility, fatigue analysis results of two ship types, performed by the developed program, were compared with those of a classification society, and it was found that both are identical. Full-spectral fatigue analysis considering hydroelasticity effects would be developed in further studies by including flexible-body ship motion analysis results and by supplementing the developed program with a wide-band fatigue damage model.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.27 no.4
• /
• pp.143-150
• /
• 1999

The purpose of this research is to find som ways to model tree forms more efficiently in reference with surveying structural data and handling parameters in plant Editor of AccuRender, the AutoCAD-based rendering software adopting the procedural plant modeling technique. In case of modelling a new tree, because it is efficient to modify an existing tree data as a template, we attempted to classify 81 species' data from existing plant library including conifers and deciduous tree. According to the qualitative characteristics and quantitative parameters of geometrical and branching structure, 8 types of tree form were classified with factor and cluster analysis. Some critical aspects found in the distributions of standardized scores of parameters in each type were discussed for explaining the tree forms intuitively. For adaptability of the resulted classification and typical parameters, 10 species of tree were measured and modelled, and proved to be very similar to the real structures of tree forms. CG or CAD-based plant modelling technique would be recommended not only as a presentation tool but for planting design, landscape simulation and assessment.

• Journal of KIISE:Software and Applications
• /
• v.28 no.11
• /
• pp.813-821
• /
• 2001

While recognition of isolated units of writing, such as a character or a word, has been extensively studied, emphasis on the segmentation itself has been lacking. In this paper we propose an active segmentation method for handwritten Hangul address strings based on the Run-length code. A slant correction algorithm, which is considered as an important preprocessing step for the segmentation, is presented. Three fundamental candidate estimation functions are introduced to detect the clues on touching points, and the classification of touching types is attempted depending on the structural peculiarity of Hangul. Our experiments show segmentation performance of 88.2% on touching characters with minimal over-segmentation.

• Journal of KIISE:Software and Applications
• /
• v.31 no.10
• /
• pp.1259-1265
• /
• 2004

In many applications, producing labeled data is costly and time consuming while an enormous amount of unlabeled data is available with little cost. Therefore, it is natural to ask whether we can take advantage of these unlabeled data in classification teaming. In machine learning literature, the co-training method has been widely used for this purpose. However, the current co-training method requires the entire features to be split into two independent sets. Therefore, in this paper, we improved the current co-training method in a number of ways, and proposed a new co-training method which do not need the feature split. Experimental results show that our proposed method can significantly improve the performance of the current co-training algorithm.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.11
• /
• pp.5530-5546
• /
• 2016

Automatic analysis and understanding of human activities is a challenging task in computer vision, especially for the surveillance scenarios which typically contains crowds, complex motions and occlusions. To address these issues, a Bag-of-words representation of videos is developed by leveraging information including crowd positions, motion directions and velocities. We infer the crowd activity in a motion field using Category Constrained Correlated Topic Model (CC-CTM) with latent topics. We represent each video by a mixture of learned motion patterns, and predict the associated activity by training a SVM classifier. The experiment dataset we constructed are from Crowd_PETS09 bench dataset and UCF_Crowds dataset, including 2000 documents. Experimental results demonstrate that accuracy reaches 90%, and the proposed approach outperforms the state-of-the-arts by a large margin.

• Journal of Korean Society for Geospatial Information Science
• /
• v.1 no.1 s.1
• /
• pp.153-158
• /
• 1993

The direction of land use for the subject district (Zuksan-myun, Anseong-gun, Kyeonggi-do) were analyzed as the basic model of new rural land management system connected to production and living. General land use planning was presented by land suitability classification which was applied to geographical information system(software ARC/INFO). The course of analysis of land suitability using the geographical information system were generalized and the results of analysis for paddy and upland fields and settlement were presented as 5 criteria of the suitability rank. It was found out that the analysis of land suitability is able to use as primary data of rural land use planning.

• Journal of Information Management
• /
• v.32 no.2
• /
• pp.40-53
• /
• 2001

The purpose of this study is to explore the current situation and problems in standardizing knowledge information and information technology in the KISTI, and collect the basic data necessary for establishing the national information infrastructure. Through a survey and interviews with nine departments of the Information Services Sector of the KISTI, data were collected regarding the management and exchange of information and the operation of communication networks. According to the analysis of fifty KISTI databases and their operation, tasks to be first standardized include Classification of ST resources, Metadata, Date modeling, Data format, DBMS, ST terminology, and Retrieval protocol.