• Journal of Korea Multimedia Society
• /
• v.12 no.3
• /
• pp.383-396
• /
• 2009

Bone age assessment has been widely used in pediatrics to identify endocrine problems of children. Since the number of trained doctors is far less than the demands, there has been numerous requests for automatic estimation of bone age. Therefore, in this paper, we propose an automatic bone age assessment method that utilizes pattern classification techniques. The proposed method consists of three modules; a finger segmentation module, a normalized shape model generation module and a bone age estimation module. The finger segmentation module segments fingers and epiphyseal regions by means of various image processing algorithms. The shape model abstraction module employ ASM to improves the accuracy of feature extraction for bone age estimation. In addition, SVM is used for estimation of bone age. Features for the estimation include the length of bone and the ratios of bone length. We evaluated the performance of the proposed method through statistical analysis by comparing the bone age assessment results by clinical experts and the proposed automatic method. Through the experimental results, the mean error of the assessment was 0.679 year, which was better than the average error acceptable in clinical practice.

• Journal of KIISE:Computing Practices and Letters
• /
• v.9 no.5
• /
• pp.551-559
• /
• 2003

This paper describes the data model and software development of XWS, an XWEET Web-wrapper System for generation wrapper program. To access information from various information sources, one has to convert and integrate source data into the same data model. XWS is developed as a part of XWEET project. We have implemented the XWS system using the Perl programming language stressing efficiency and ease-of-use. XWS has a few distinct features. First, data model and operator used for extracting information from HTML support a unified model of different views of HTML document. Second, it provides a user-friendly interface program to enable wrapper programmer to generate wrapper easily Third, XWS use the high-level script language designed by object-oriented methodology. In this paper, we also present the detail demonstration where it is useful for extracting article information from DBLP site.

• Journal of KIISE:Software and Applications
• /
• v.34 no.10
• /
• pp.889-899
• /
• 2007

In proteomics, recent advancements In mass spectrometry technology and in protein extraction and separation technology made high-throughput analysis possible. This leads to thousands to hundreds of thousands of MS/MS spectra per single LC-MS/MS experiment. Such a large amount of data creates significant computational challenges and therefore effective data analysis methods that make efficient use of computational resources and, at the same time, provide more peptide identifications are in great need. Here, SIFTER system is designed to avoid inefficient processing of shotgun proteomic data. SIFTER provides software tools that can improve throughput of mass spectrometry-based peptide identification by filtering out poor-quality tandem mass spectra and estimating a Peptide charge state prior to applying analysis algorithms. SIFTER tools characterize and assess spectral features and thus significantly reduce the computation time and false positive rates by localizing spectra that lead to wrong identification prior to full-blown analysis. SIFTER enables fast and in-depth interpretation of tandem mass spectra.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.40 no.8
• /
• pp.1655-1666
• /
• 2015

In this paper, user-position estimation method is proposed by using a single camera for both indoor and outdoor environments. Conventionally, the GPS of RF-based estimation methods have been widely studied in the literature for outdoor and indoor environments, respectively. Each method is useful only for indoor or outdoor environment. In this context, this study adopts a vision-based approach which can be commonly applicable to both environments. Since the distance or position cannot be extracted from a single still image, the reference images pro-stored in image database are used to identify the current position from the single still image captured by a single camera. The reference image is tagged with its captured position. To find the reference image which is the most similar to the current image, the SURF algorithm is used for feature extraction. The outliers in extracted features are discarded by using RANSAC algorithm. The performance of the proposed method is evaluated for two buildings and their outsides for both indoor and outdoor environments, respectively.

• The Journal of the Korea Contents Association
• /
• v.10 no.4
• /
• pp.19-27
• /
• 2010

In recent years, high-throughput genome-wide RNA interference screening is emerging as an essential tool to biologists in understanding complex cellular processes. The manual analysis of the large number of images produced in each study spends much time and the labor. Hence, automatic cellular image analysis becomes an urgent need, where segmentation is the first and one of the most important steps. However, those factors such as the region overlapping, a variety of shapes, and non-uniform local characteristics of cellular images become obstacles to efficient cell segmentation. To avoid the problem, a new watershed-based cell segmentation algorithm using a localized segmentation method and a feature vector is proposed in this paper. Localized approach in segmentation resolves the problems caused by a variety of shapes and non-uniform characteristics. In addition, the poor performance of segmentation in overlapped regions can be improved by taking advantage of a feature vector whose component features complement each other. Simulation results show that the proposed method improves the segmentation performance compared to the method in Cellprofiler.

• The Korean Journal of Pesticide Science
• /
• v.2 no.1
• /
• pp.40-45
• /
• 1998

An antagonistic fungus inhibiting growth of various phytopathogenic fungi was isolated from greenhouse soils and identified. Mophological features of fruiting structures on potato dextrose agar and colorless globose to ovate heavy walled hyaline cells from the vegetative mycelium grown on MY20 agar indicate that this antagonist is Aspergillus terreus. The antagonistic activity is due to the production of antifungal compounds. An antifungal compound was purified from its culture filtrate using chloroform extraction, column chromatography, and thin layer chromatography. The purified antibiotic was effective to various phytopathogenic fungi and identified as butyrolactone I. $ED_{50}$ values measured by petri-plate assay through effective dosage(ED)-probit analysis were 9.7, 13.7, 23.3, 42.6 and 102.7 ppm on Botrytis cinerea, Rhizoctonia solani, Pythium ultimum, Phytophthora capsici, and Fusarium oxysporum, respectively.

• Rapid Communication in Photoscience
• /
• v.3 no.1
• /
• pp.16-19
• /
• 2014

Anodic self-organized titania nanotube (TNT) arrays have a great potential as efficient electron-transport materials for dye-sensitized solar cells (DSSC). Herewith we report the photovoltaic and kinetic investigations for a series of heteroleptic ruthenium complexes (RD16-RD18) sensitized on TNT films for DSSC applications. We found that the RD16 device had an enhanced short-circuit current density ($J_{SC}/mAcm^{-2}=15.0$) and an efficiency of power conversion (${\eta}=7.2%$) greater than that of a N719 device (${\eta}=7.1%$) due to the increasing light-harvesting and the broadened spectral features with thiophene-based ligands. However, the device made of RD17 (adding one more hexyl chain) showed smaller $J_{SC}(14.1mAcm^{-2})$ and poorer ${\eta}(6.8%)$ compare to those of RD16 due to smaller amount of dye-loading and less efficient electron injection for the RD17 device than for the RD16 device. For the RD18 dye (adding one more thiophene unit and one more hexyl chain), we found that the device showed even lower $J_{SC}(13.2mAcm^{-2}) $ that led to a poorest device performance (${\eta}=6.2%$) for the RD18 device. These results are against to those obtained from the same dyes sensitized on $TiO_2$ nanoparticle films and they can be rationalized according to the electron transport kinetics measured using the methods of charge extraction and transient photovoltage decays.

• The Korean Journal of Applied Statistics
• /
• v.29 no.5
• /
• pp.859-872
• /
• 2016

Land cover classification is an important tool for preventing natural disasters, collecting environmental information, and monitoring natural resources. Hyperspectral imaging is widely used for this task thanks to sufficient spectral information. However, the curse of dimensionality, spatiotemporal variability, and lack of labeled data make it difficult to classify the land cover correctly. We propose a novel classification framework for land cover classification of hyperspectral data based on convolutional neural networks. The proposed framework naturally incorporates full spectral features with the information from neighboring pixels and has advantages over existing methods that require additional feature extraction or pre-processing steps. Empirical evaluation results show that the proposed framework provides good generalization power with classification accuracies better than (or comparable to) the most advanced existing classifiers.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2014.10a
• /
• pp.557-561
• /
• 2014

In this work, Altair Engineering's vibroacoustic modeling approach is used to simulate the acoustic signature of a simplified automobile in a wind tunnel. The modeling approach relies on a two step procedure involving simulation and extraction of acoustic sources using a high fidelity Computational Fluid Dynamics (CFD) simulation followed by propagation of the acoustic energy within the structure and passenger compartment using a structural dynamics solver. The tools necessary to complete this process are contained within Altair's HyperWorks CAE software suite. The CFD simulations are performed using AcuSolve and the structural simulations are performed using OptiStruct. This vibroacoustics simulation methodology relies on calculation of the acoustic sources from the flow solution computed by AcuSolve. The sources are based on Lighthill's analogy and are sampled directly on the acoustic mesh. Once the acoustic sources have been computed, they are transformed into the frequency domain using a Fast Fourier Transform (FFT) with advanced sampling and are subsequently used in the structural acoustics model. Although this approach does require the CFD solver to have knowledge of the acoustic simulation domain a priori, it avoids modeling errors introduced by evaluation of the acoustic source terms using dissimilar meshes and numerical methods. The aforementioned modeling approach is demonstrated on the Hyundai Simplified Model (HSM) geometry in this work. This geometry contains flow features that are representative of the dominant noise sources in a typical automobile design; namely vortex shedding from the passenger compartment A-pillar and bluff body shedding from the side view mirrors. The geometry also contains a thick poroelastic material on the interior that acts to reduce the acoustic noise. This material is modeled using a Biot material formulation during the structural acoustic simulation. Successful prediction of the acoustic noise within the HSM geometry serves to validate the vibroacoustic modeling approach for automotive applications.

• Journal of KIISE:Computing Practices and Letters
• /
• v.8 no.6
• /
• pp.705-716
• /
• 2002

This paper proposes a method which visualizes MRI head data in 3 dimensions with direct volume rendering. Though surface rendering is usually used for MRI data visualization, it has some limits of displaying little speckles because it loses the information of the speckles in the surfaces while acquiring the information. Direct volume rendering has ability of displaying little speckles, but it doesn't treat MRI data because of the data features of MRI. In this paper, we try to visualize MRI head data in 3 dimensions as follows. First, we separate the brain region from the head region of MRI head data, next increase the pixel level of the brain region, then combine the brain region with the increased pixel level and the head region without brain region, last visualizes the combined MRI head data with direct volume rendering. We segment the brain region from head region based on histogram threshold, morphology operations and snakes algorithm. The proposed segmentation method shows 91~95% similarity with a hand segmentation. The method rather clearly visualizes the organs of the head in 3 dimensions.