• The KIPS Transactions:PartB
• /
• v.10B no.4
• /
• pp.443-450
• /
• 2003

In this paper, we proposed the novel hierarchical algorithm for the recognition of English calling cards that processes multiresolution images of calling cards hierarchically to extract individual characters and recognizes the extracted characters by using the enhanced neural network method. The hierarchical recognition algorithm generates multiresolution images of calling cards, and each processing step in the algorithm selects and processes the image with suitable resolution for lower processing overhead and improved output. That is, first, the image of 1/3 times resolution, to which the horizontal smearing method is applied, is used to extract the areas including only characters from the calling card image, and next, by applying the vertical smearing and the contour tracking masking, the image of a half time resolution is used to extract individual characters from the character string areas. Lastly, the original image is used in the recognition step, because the image includes the morphological information of characters accurately. And for the recognition of characters with diverse font types and various sizes, the enhanced RBF network that improves the middle layer based on the ART1 was proposed and applied. The results of experiments on a large number of calling card images showed that the proposed algorithm is greatly improved in the performance of character extraction and recognition compared with the traditional recognition algorithms.

• Journal of the Korean Wood Science and Technology
• /
• v.41 no.4
• /
• pp.302-309
• /
• 2013

The highly porous cellulosic aerogels were prepared by freeze-drying method using sodium hydroxide-urea aqueous solution in the process of dissolution, gelation, regeneration and organic solvent substitution. The structural characteristics of porous aerogel were analyzed using scanning electron microscopy and nitrogen adsorption apparatus. As a result, the dissolving pulp was completely dissolved, but filter papers and holocellulose were divided into two layers (dissolved and undissolved parts) in the process of centrifugation. The structure of aerogel from dissolved pulp showed porous pores in the surface and net-shaped network in the inner part. Aerogels from filter paper and holocellulose had the condensed porous network surface and the open-pore nano-fibril network inner structure. Undissolved form of fibers was observed in the aqueous solution of aerogel from holocellulose. The BET value ($S_{BET}$) of aerogel from dissolved pulp was ranged in 260~326 $m^2/g$, and it was decreased with the increase of concentration. Whereas, the $S_{BET}$ value of aerogel from filter paper (198~418 $m^2/g$) was increased with the increase of concentration. The $S_{BET}$ value of aerogel from holocellulose were 137 $m^2/g$ at 2% (w/w) of cellulose, and it was increased to maximum 401 $m^2/g$ at 4% (w/w) of cellulose. Then, it was decreased at 5% (w/w) of cellulose.

• KOREAN JOURNAL OF CROP SCIENCE
• /
• v.20
• /
• pp.27-62
• /
• 1975

These studies were aimed at clarifying whether or not selecting agronomic characteristics of hybrid-rice populations under greenhouse conditions was useful. The selection of two quantitative characteristics, culm length and heading date, and two qualitative characteristics, shattering and awnedness, was very effective, but the flag leaf length and the exsertion of panicle was not consistent, varying from one cross to another.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.10c
• /
• pp.202-204
• /
• 2001

가상 세계의 공유 개념은, 특히 사용자들이 인터넷 같이 대규모 네트웍을 통해 지역적으로 분산된 경우는 복제가 수용할 수 있는 상호작용 성능을 제공하기 때문에 각 사용자의 사이트에 정보를 복제함으로써 확장된다. 그러나, 다수의 동시 갱신은 replicas간의 일관되지 않은 뷰를 일으키게 될 것이다. 따라서, 동시성 제어가 복제자들간에 일관된 상태를 유지하도록 하기 위한 중요한 요소가 된다. 우리는 단지 대상 객체의 주변에 있는 사용자들만이 소유권 요청을 다중 전송하게 하는 확장성 있는 예측기반 동시성 제어 스킴을 제안했었다. 이 작업에서, 우리는 모든 사용자들이 동일한 속도론 가지고 가상 세계를 이동한다고 가정했다. 이것은, 그러나, 좀더 사실성을 더하기 위해 사용자가 가상 세계와 상호작용을 할 매 그들의 이동속도를 변경하도록 하는 네트웍 게임같은 네트웍 가상 환경에서는 너무 common 하다. 본 논문은 다양한 속도를 가진 사자를 지원하기 위한 확장을 제안한다. 확장된 스킴은 다른 속도의 수만큼의 다중 Entity Radii를 가지며 각 속도를 가진 사용자에게 분리된 큐를 할당한다. 각 큐는 다음 소유자 후보를 예측하기 위해 동시에 예측을 수행하고 선택된 후보들간에서 최소의 Predicted Collision Time을 가지는 최종 후보자가 선택된다. 이는 사용자의 속도에 기반을 둔 적절한 Entity Radius를 사용함으로써 소유권의 timely advanced transfer과, 다른 이돔 속도와 latency를 가지는 사용자들 간의 간섭을 줄임으로써 공정(공평)한 소유권 양도, 그리고 불필요한 소유권 전송을 줄임으로써 놓은 예측 정확도를 제공한다.성을 지닌 AMMQL 학습법은 로봇축구와 같이 끊임없이 실시간적으로 변화가 일어나는 다중 에이전트 환경에서 특히 높은 효과를 볼 수 있다. 본 논문에서는 AMMQL 학습방법의 개념을 소개하고, 로봇축구 에이전트의 동적 위치 결정을 위한 학습에 어떻게 이 학습방법을 적용할 수 있는지 세부 설계를 제시한다.다.으로서 hemicellulose구조가 polyuronic acid의 형태인 것으로 사료된다. 추출획분의 구성단당은 여러 곡물연구의 보고와 유사하게 glucose, arabinose, xylose 함량이 대체로 높게 나타났다. 점미가 수가용성분에서 goucose대비 용출함량이 고르게 나타나는 경향을 보였고 흑미는 알칼리가용분에서 glucose가 상당량(0.68%) 포함되고 있음을 보여주었고 arabinose(0.68%), xylose(0.05%)도 다른 종류에 비해서 다량 함유한 것으로 나타났다. 흑미는 총식이섬유 함량이 높고 pectic substances, hemicellulose, uronic acid 함량이 높아서 콜레스테롤 저하 등의 효과가 기대되며 고섬유식품으로서 조리 특성 연구가 필요한 것으로 사료된다.리하였다. 얻어진 소견(所見)은 다음과 같았다. 1. 모년령(母年齡), 임신회수(姙娠回數), 임신기간(姙娠其間), 출산시체중등(出産時體重等)의 제요인(諸要因)은 주산기사망(周産基死亡)에 대(對)하여 통계적(統計的)으로 유의(有意)한 영향을 미치고 있어 $25{\sim}29$세(歲)의 연령군에서, 2번째 임신과 2번째의 출산에서 그리고 만삭의 임신 기간에, 출산시체중(出産時體重) $3.50{\sim}3.99kg$사이의 아

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.125-140
• /
• 2018

As one of the processes in the manufacturing industry, quality inspection inspects the intermediate products or final products to separate the good-quality goods that meet the quality management standard and the defective goods that do not. The manual inspection of quality in a mass production system may result in low consistency and efficiency. Therefore, the quality inspection of mass-produced products involves automatic checking and classifying by the machines in many processes. Although there are many preceding studies on improving or optimizing the process using the data generated in the production process, there have been many constraints with regard to actual implementation due to the technical limitations of processing a large volume of data in real time. The recent research studies on big data have improved the data processing technology and enabled collecting, processing, and analyzing process data in real time. This paper aims to propose the process and details of applying big data for quality inspection and examine the applicability of the proposed method to the dairy industry. We review the previous studies and propose a big data analysis procedure that is applicable to the manufacturing sector. To assess the feasibility of the proposed method, we applied two methods to one of the quality inspection processes in the dairy industry: convolutional neural network and random forest. We collected, processed, and analyzed the images of caps and straws in real time, and then determined whether the products were defective or not. The result confirmed that there was a drastic increase in classification accuracy compared to the quality inspection performed in the past.

• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.5
• /
• pp.345-349
• /
• 2009

Social bookmarking systems are a typical web 2.0 service based on folksonomy, providing the platform for storing and sharing bookmarking information. Spammers in social bookmarking systems denote the users who abuse the system for their own interests in an improper way. They can make the entire resources in social bookmarking systems useless by posting lots of wrong information. Hence, it is important to detect spammers as early as possible and protect social bookmarking systems from their attack. In this paper, we applied a diverse set of machine learning approaches, i.e., decision tables, decision trees (ID3), $na{\ddot{i}}ve$ Bayes classifiers, TAN (tree-augment $na{\ddot{i}}ve$ Bayes) classifiers, and artificial neural networks to this task. In our experiments, $na{\ddot{i}}ve$ Bayes classifiers performed significantly better than other methods with respect to the AUC (area under the ROC curve) score as veil as the model building time. Plausible explanations for this result are as follows. First, $na{\ddot{i}}ve$> Bayes classifiers art known to usually perform better than decision trees in terms of the AUC score. Second, the spammer detection problem in our experiments is likely to be linearly separable.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.52 no.10
• /
• pp.21-29
• /
• 2015

For higher reliability against component failures in railroad trains with many electronic sensors and actuators, a distributed control system with which all electronic components are connected via a network is being considered. This paper compares and analyzes various topologies of Ethernet network for a railroad train in the aspects of (1) failure recovery, (2) the number of ports per device, (3) the number of cable connections between vehicles, and (4) performance. Especially, the unique characteristic of a train system that the number of vehicles changes is considered through analysis. Various combinations of in- and inter-vehicle topologies are considered. In addition, we introduce a hybrid of star and daisy-chain topology for inter-vehicle connection when the maximum number of inter-vehicle connections is limited to reduce possible failures of inter-vehicle connections. Simulation results show performance comparison between different topology combinations; the hybrid topology is shown to enhance delay performance even with a highly limited number of inter-vehicle connections.

• Journal of the Korean Geotechnical Society
• /
• v.27 no.5
• /
• pp.85-92
• /
• 2011

The dynamic behavior of piles becomes very complex due to soil-pile dynamic interaction, soil non-linearity, resonance phenomena of soil-pile system and so on. Therefore, the proper numerical simulation of the pile behavior needs much effort and calculation time. In this research, a new modeling method, which can be applied to the conventional finite difference analysis program FLAC 3D, was developed to reduce the calculation time. The soil domain in this method is divided into a near-field region and a far-field region, which is not influenced by the soil-pile dynamic interaction. Then, the ground motion of the far-field is applied to the boundaries of the near-field instead of modeling the far-field region as finite meshes. In addition, the soil non-linearity behavior is modeled by using the hysteretic damping model, which determines the soil tangent modulus as a function of shear strain and the interface element was applied to simulate the separation and slip between the soil and pile. The proposed method reduced the calculation time by as much as one third compared with a usual modeling method and maintained the accuracy of the calculated results. The calculated results by the proposed method showed a good agreement with the prototype pile behavior, which was obtained by applying a similitude law to the 1-g shaking table test results.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.22 no.5
• /
• pp.241-265
• /
• 1989

The purpose of present study is to describe the taxonomy and early life history of Korean pomfrets, genus Pampus. Egg diameter varied from $0.83\~1.27mm\;and\;1.05\~1.41mm$ in P. argenteus and P. echinogaster respectively. Larvae of P. echinogaster of 16.5 mm in total length has full number of fin rays and enter the juvenile period. The egg is matured at 185.0 mm in total length in P. argenteus and at 95.5 mm in total length in P. echinogaster. P. echinogaster is ossified from jaw bone at $8.5\~10.0mm$ in to1 length and followed by ossification of vertebrae and visceral skeleton. P. argenteus and P. echinogaster are distinguished by the morphology of jaw, fin and isthmus. In P. argenteus, wavy ridges extends over the upper part of pectoral fin and gill membrane is broadly united to isthmus, while in P. echinogaster the ridges are limited to the base of pectoral fin and gill membrane reaching isthmus. These Pampus spp. are easily distinguished based on internal features, such as number of vertebrae, gill raker, pyloric caeca, epipleural bone, esophageal sat and length of intestine.

• The Journal of Engineering Geology
• /
• v.21 no.2
• /
• pp.107-116
• /
• 2011

Groundwater level data and flowrate data were collected by considering the distance between the groundwater-level gauge station and flowrate gauge station (< 10 km) in the Geum River watershed, Baseflow was separated from the collected flowrate data using the WHAT SYSTEM, which is a Web- and GIS-based tool developed for hydrological applications, Correlation analysis was performed for the separated baseflow and groundwater-level data collected from a site close to the flowrate gauge station, Twenty regions were selected and data sets were collected from 2002 to 2008. Twelve regions yielded a correlation coefficient of > 0.5, When the data sets were analyzed for each year for all 20 regions, we obtained a correlation coefficient of 0.5-0.6 for 8 cases, 0.6-0.7 for 5 cases, and > 0.7 for 12 cases. For individual regions, the correlation coefficient varied from year to year. There was a tendency toward weak correlation in the case of drought or flood, Therefore, under normal conditions (i.e., neither flood nor severe drought), it is possible to estimate the baseflow from nearby groundwater-level data for regions with a high correlation coefficient.