• Journal of radiological science and technology
• /
• v.30 no.4
• /
• pp.419-426
• /
• 2007

In this paper, the input pattern vectors are extracted and the learning algorithms is designed to recognize taste(bitter, sweet, sour and salty) pattern vectors. The signal intensity of taste are used to compose the input pattern vectors. The SOM(Self Organizing Maps) algorithm for taste pattern recognition is used to learn initial reference vectors and the ot-star learning algorithm is used to determine the class of the output neurons of the sunclass layer. The weights of the proposed algorithm which is between the input layer and the subclass layer can be learned to determine initial reference vectors by using SOM algorithm and to learn reference vectors by using LVQ(Learning Vector Quantization) algorithm. The pattern vectors are classified into subclasses by neurons in the subclass layer, and the weights between subclass layer and output layer are learned to classify the classified subclass, which is enclosed a class. To classify the pattern vectors, the proposed algorithm is simulated with ones of the conventional LVQ, and it is confirmed that the proposed learning method is more successful classification than the conventional LVQ.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.5 s.43
• /
• pp.95-103
• /
• 2006

In this paper, the input pattern vectors are extracted and the learning algorithms is designed to recognize EEG pattern vectors. The frequency and amplitude of alpha rhythms and beta rhythms are used to compose the input pattern vectors. And the algorithm for EEG pattern recognition is used SOM to learn initial reference vectors and out-star learning algorithm to determine the class of the output neurons of the subclass layer. The weights of the proposed algorithm which is between the input layer and the subclass layer can be learned to determine initial reference vectors by using SOM algorithm and to learn reference vectors by using LVQ algorithm, and pattern vectors is classified into subclasses by neurons which is being in the subclass layer, and the weights between subclass layer and output layer is learned to classify the classified subclass, which is enclosed a class. To classify the pattern vectors of EEG, the proposed algorithm is simulated with ones of the conventional LVQ, and it was a confirmation that the proposed learning method is more successful classification than the conventional LVQ.

• Korean Journal of Remote Sensing
• /
• v.39 no.3
• /
• pp.339-353
• /
• 2023

As the availability of high-resolution satellite imagery increases, improvement of positioning accuracy of satellite images is required. The importance of orthorectified images is also increasing, which removes relief displacement and establishes true localization of man-made structures. In this paper, we performed automated extraction of building rooftops and total building areas within original satellite images using the existing building height database. We relocated the rooftop sin their true position and generated an orthorectified building layer. The extracted total building areas were used to blank out building areas and generate true orthographic non-building layer. A final orthorectified image was provided by overlapping the building layer and non-building layer.We tested the proposed method with KOMPSAT-3 and KOMPSAT-3A satellite images and verified the results by overlapping with a digital topographical map. Test results showed that orthorectified building layers were generated with a position error of 0.4m.Through the proposed method, the feasibility of automated true orthoimage generation within dense urban areas was confirmed.

• Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference
• /
• 2009.04a
• /
• pp.11-13
• /
• 2009

This research is renewal way to get latest digital map, We presented techniques to convert to GIS data for update digital map to utilize completion drawing of CAD data that is used construction and SOC construction. Conversion process is consisted of layer extraction, object transform, coordinate transform, format transform. GIS data that is changed via each process from CAD data can update digital map.

• Toxicological Research
• /
• v.13 no.1_2
• /
• pp.161-165
• /
• 1997

This study was conducted to investigate the skin irritation by transdermal administration of the three dyes. These dyes were originated from onion by using extraction method. By the order of extraction from onion, A-dye was obtained from onion by using water at 90-100$\circ$C. B-dye was extracted from A-dye with ethylacetate. After ethylacetate extraction from A-dye, the lower layer named as C-dye. Twenty-four New Zealand white rabbits were divided into three groups. The each groups was consisted of two subgroups according to high dose (extracted dyes) and low dose (the 100-fold dilutions of A-, Band C-dye). In primary skin irritation test of male New Zealand White rabbits, body temperature and weights were not significantly changed and blood cells were positioned in normal blood cell ranges of health rabbits. Primary irritation index was "0" in the test and control sites of all animals used in this study. By the results obtained in the present test, all dyes were evaluated as a non-irritant on the basis of the criteria of Draize.of Draize.

• Bulletin of the Korean Chemical Society
• /
• v.35 no.1
• /
• pp.45-50
• /
• 2014

A fully packed capillary electrochromatographic (CEC) microchip with an on-column micro-solid phase extraction (SPE) bed for the preconcentration and separation of organic analytes was prepared. A linear microchannel with monodisperse colloidal silica packing was formed on a cyclic olefinic copolymer microchip with two reservoirs on both ends. Silver-cemented silica packing frit structure was formed at the entrance of the microchannel by electroless plating treatment as a base layer. A gold coating was formed on it by reducing $Au^{3+}$ to gold with hydroxylamine. Finally micro-SPE bed was formed by self-assembly adsorption of 1-dodecanethiol on it. Micro-SPE beds were about 100-150 ${\mu}m$ long. Approximately $10^3$ fold sensitivity enhancements for Sulforhodamine B, and Fluorescein in nM concentration levels were possible with 80 s preconcentration. Basic extraction characteristics were studied.

• Journal of Dairy Science and Biotechnology
• /
• v.17 no.1
• /
• pp.11-15
• /
• 1999

The objective of this study was to extract and purity the antibacterial agent from the fermented milk with Lactobacillus helveticus CH-1. The extraction and purification of antibacterial agent from the Lb. helveticus fermented milk were carried out by methanol extraction, acetone extraction, Sephadex G-200 gel filteration and thin layer chromatography and the results were as followings. The antibacterial activity of methanol-acetone extraction showed antibacterial activity against test organisms, B. subtilis, E. coli, Pseu. fluorescens, Sal. typhimurium, Shi. flexneri, and Sta. aureus. Sephadex G-200 gel chromatography showed only antibacterial activity from 33 to 37th fractions of 60 fractions. The agent purified from TLC plate confirmed the antibacterial activity by the means of bioautography.

• Journal of Ginseng Research
• /
• v.45 no.1
• /
• pp.1-21
• /
• 2021

Panax species have gained numerous attentions because of their various biological effects on cardiovascular, kidney, reproductive diseases known for a long time. Recently, advanced analytical methods including thin layer chromatography, high-performance thin layer chromatography, gas chromatography, high-performance liquid chromatography, ultra-high performance liquid chromatography with tandem ultraviolet, diode array detector, evaporative light scattering detector, and mass detector, two-dimensional high-performance liquid chromatography, high speed counter-current chromatography, high speed centrifugal partition chromatography, micellar electrokinetic chromatography, high-performance anion-exchange chromatography, ambient ionization mass spectrometry, molecularly imprinted polymer, enzyme immunoassay, 1H-NMR, and infrared spectroscopy have been used to identify and evaluate chemical constituents in Panax species. Moreover, Soxhlet extraction, heat reflux extraction, ultrasonic extraction, solid phase extraction, microwave-assisted extraction, pressurized liquid extraction, enzyme-assisted extraction, acceleration solvent extraction, matrix solid phase dispersion extraction, and pulsed electric field are discussed. In this review, a total of 219 articles published from 1980 to 2018 are investigated. Panax species including P. notoginseng, P. quinquefolius, sand P. ginseng in the raw and processed forms from different parts, geographical origins, and growing times are studied. Furthermore, the potential biomarkers are screened through the previous articles. It is expected that the review can provide a fundamental for further studies.

• Current Photovoltaic Research
• /
• v.1 no.1
• /
• pp.59-62
• /
• 2013

We investigated the effects of ZnO thin film deposition methods on the performance of inverted polymer solar cells with a structure of ITO/ZnO/P3HT:PCBM/MoO3/Al. The ZnO thin films were deposited by various methods (spin coating of nanoparticles, sol-gel process, atomic layer deposition) and their morphology was analyzed by atomic force microscopy (AFM). The device with ZnO nanoparticle thin films showed the highest power conversion efficiency of 3 % with low series resistance and high shunt resistance. The superior performance of the device with the ZnO nanoparticle layer is attributed to better electron extraction capability.

• Journal of Multimedia Information System
• /
• v.9 no.3
• /
• pp.177-182
• /
• 2022

A deep neural network (DNN) includes variables whose values keep on changing with the training process until it reaches the final point of convergence. These variables are the co-efficient of a polynomial expression to relate to the feature extraction process. In general, DNNs work in multiple 'dimensions' depending upon the number of channels and batches accounted for training. However, after the execution of feature extraction and before entering the SoftMax or other classifier, there is a conversion of features from multiple N-dimensions to a single vector form, where 'N' represents the number of activation channels. This usually happens in a Fully connected layer (FCL) or a dense layer. This reduced 2D feature is the subject of study for our analysis. For this, we have used the FCL, so the trained weights of this FCL will be used for the weight-class correlation analysis. The popular DNN models selected for our study are ResNet-101, VGG-19, and GoogleNet. These models' weights are directly used for fine-tuning (with all trained weights initially transferred) and scratch trained (with no weights transferred). Then the comparison is done by plotting the graph of feature distribution and the final FCL weights.