• International Journal of Contents
• /
• v.8 no.1
• /
• pp.39-46
• /
• 2012

Character segmentation is a preprocessing step in many offline handwriting recognition systems. In this paper, Chinese characters are categorized into seven different structures. In each structure, the character size with the range of variations is estimated considering typical handwritten samples. The component removal and merge criteria are presented to remove punctuation symbols or to merge small components which are part of a character. Finally, the criteria for segmenting the adjacent characters concerning each other or overlapped are proposed.

• Proceedings of the Korean Information Science Society Conference
• /
• 2002.10d
• /
• pp.571-573
• /
• 2002

음성인식과 온라인 필기인식에서 우수한 성능을 보이는 은닉 마르코프(HMM)의 HMM의 구조는 휴리스틱 한 방법에 의해 결정되는 것이 일반적이기 때문에 최적의 모델을 선택하는데 어려움이 있다. 이에 본 논문에서는 HMM의 구조를 체계적인 방법으로 정함과 동시에 변별력의 단점을 개선 할 수 있는 방법으로 Anti-likelihood를 이용한 모델간의 변별력을 살펴보고 최적의 모델 선택 기준인 BIC와의 결합하여, 체계적이고 효율적인 최적 모델 선택이 가능한 방법론에 대해 연구하고 필기데이터에 대해 검증한 결과, 기존의 방법보다 파라미터의 수는 감소되고 인식률이 향상됨을 알 수 있다.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2004.05a
• /
• pp.285-288
• /
• 2004

온라인 필기체 한자는 동일인의 동일 문자조차도 회수, 획순 및 형태의 변화가 다양할 뿐만 아니라 인식 대상이 방대하여 인식이 매우 어렵다. 또한 한자는 기본 자소의 조합에 의한 글자가 아닌 각각의 글자가 독립적으로 이루어져 있어 연속된 획들 간의 관련도를 파악하기 어렵고 획수도 1획에서 28획까지 다양하게 분포를 한다. 따라서 본 연구에서는 대분류 단계로 시작획 비교를 하고 이어진 세분류 단계에서 문자의 특징으로 방향코드와 특이점을 추출해내고 획수를 고려하여 DP 정합을 하는 2단계 인식 시스템을 제안하였다. 이로써 최적의 속도로 입력한 문자를 찾아낼 수 있도록 하였다.

• Human Ecology Research
• /
• v.52 no.1
• /
• pp.21-31
• /
• 2014

The purpose of this study was to analyze the types and meanings of calligraphy presented in modern fashion design. Calligraphy refers to beautiful handwriting or fine penmanship in the West, and handwriting with brushstrokes in the East. The expression patterns being used at present can be divided into three categories. Legible calligraphy is focused on readability more than embellishment. Decorative calligraphy places its importance on decoration at the expense of practicality. The third type, harmonious calligraphy, pursues decorativeness and legibility at the same time. Each of these types of calligraphy is expressed in modern fashion with its own purpose: calligraphy for conveying emotional messages, calligraphy as a special brand image, and calligraphy as an expression of formativeness. The first, calligraphy for conveying emotional messages, is used with characters that are familiar to the public. Calligraphy of this type delivers messages confined emotionally to the conscious world, harmonizing calligraphy with words, or expressing readability filled with purity and delight. Second, calligraphy as a special brand image refers to transmitting a distinctive brand image from other companies through employment of a design motive or pattern by expressing the brand logos or names of designers. Third, calligraphy as a expression of formativeness has the function of shaping expressions as motives or patterns, avoiding meanings of words or phrases. It can be represented by the abbreviation or modification of words, or arranging words in different shapes, harmonizing the words with the clothing construction and atmosphere of the other images.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.23 no.3
• /
• pp.167-173
• /
• 2023

The cash flow matching problem(CFMP) aims to minimize the initial investment by paying the total amount due for the T-year in principal and interest of bonds or bank deposits without paying the full amount in cash. Linear programming(LP) is the only known way to solve CFMP. The linear programming method is a problem that optimizes T linear functions, and it cannot be solved by handwriting, so LINGO, which is a solution to the linear programming method, is used. This paper proposes an algorithm that obtains the solution of CFMP solely by handwriting without the help of LINGO. The proposed algorithm determines the amount of bond purchases by covering payments until the previous year of the next maturity bond in the order that the maturity date falls from the longest to the short term. In addition, until the year before the maturity of the shortest maturity bond, the amount of deposit covered by the principal and interest of the bank deposit was determined. As a result of applying the proposed algorithm to two experimental data, it was shown that more accurate results can be obtained compared to the linear programming method.

• The Journal of the Convergence on Culture Technology
• /
• v.10 no.2
• /
• pp.453-458
• /
• 2024

Optical Character Recognition is the technology that recognizes text in images and converts them into digital format. Deep learning-based OCR is being used in many industries with large quantities of recorded data due to its high recognition performance. To improve medical services, deep learning-based OCR was actively introduced by the medical industry. In this paper, we discussed trends in OCR engines and medical OCR and provided a roadmap for development of medical OCR. By using natural language processing on detected text data, current medical OCR has improved its recognition performance. However, there are limits to the recognition performance, especially for non-standard handwriting and modified text. To develop advanced medical OCR, databaseization of medical data, image pre-processing, and natural language processing are necessary.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.180-185
• /
• 2003

In order to enhance dexterity in execution of robot tasks, a redundant number of degrees-of-freedom (DOF) is adopted for design of robotic mechanisms like robot arms and multi-fingered robot hands. Associated with such redundancy in the number of DOFs relative to the number of physical variables necessary and sufficient for description of a given task, an extra performance index is introduced for controlling such a redundant robot in order to avoid arising of an ill-posed problem of inverse kinematics from the task space to the joint space. This paper shows that such an ill-posedness of DOF redundancy can be resolved in a natural way by using a novel concept named “stability on a manifold”. To show this, two illustrative robot tasks 1) robotic handwriting and 2) control of an object posture via rolling contact by a multi-DOF finger are analyzed in details.

• Journal of architectural history
• /
• v.26 no.1
• /
• pp.83-92
• /
• 2017

This paper studied Lee Han-Choel's career who was known as the first Korean who had participated in several survey and repair projects of Korean historic buildings as a field engineer during the Japanese colonial era. This paper investigated documents and drawings to find out what he had done specifically in the field, and distinguished the records he had written by handwriting comparison method. In addition, the author analyzed the contents of the articles he published and clarified their significance in Korean architectural history. Through this study, the author expected to clarify Lee's historical status as the first Korean modern expert in the field of historic building conservation, and to supplement the research for the history of Korean during the Japanese colonial era.

• Proceedings of the IEEK Conference
• /
• 2001.09a
• /
• pp.421-424
• /
• 2001

개별적인 인식기를 하나의 단일 인식 시스템으로 구성하여 음성과 문자를 인식할 수 있는 공용인식시스템의 성능향상을 위해 온라인 필기에서 전역적인 정보를 추출할 수 있는 비트맵 파라미터 추출 방법을 제안하였다. 제안된 방식에서는 고속의 파라미터 추출을 위해 보간법을 이용한 재샘플링 과정 대신에 새로운 시간열을 구성하는 방식을 이용한다. 제안한 비트맵 파라미터를 본 연구실에서 개발한 음성/문자 공용인식 시스템에 적용하기 위하여 67개의 자소를 5상태 10천이 CHMM(Continuous Hidden Markov Model)모델로 구성한 다음 인식알고리즘으로서는 상태단위로 지속 시간 정보를 제어하는 OnePassDP법을 이용하였다. 실험결과, 제안한 방법을 이용한 경우, 자소인식률은 61.3%에서 85.3%로 24%의 인식률 향상을 가져왔으며, 글자인식률은 64.3%에서 82.2%로 17.9%의 인식률 향상을 가져와 제안한 방법의 유효성을 확인할 수 있었다.

• Annual Conference on Human and Language Technology
• /
• 1995.10a
• /
• pp.152-163
• /
• 1995

본 논문에서는 은닉 마르코프 모형(HMM)을 이용하여 사람이 쓴 필기의 통계적 특징을 갖는 글씨를 생성하는 방법에 대해서 기술코자 한다. 온라인 필기처럼 같이 필기 궤적을 시계열 신호로 표현하고, 그 특징을 통계적 모형의 하나인 HMM으로 표현한다. HMM은 시계열 신호에 대응하는 모형 내부 경로와 심볼열의 확률 분포를 표현하는 함수이다. 따라서 최적 경로에서 볼 수 있는 최적 출력 심볼열은 훈련 필기 데이타의 평균적 필기 특징에 해당한다. HMM이 주어졌을 때 HMM에서 최적의 패턴을 해석적으로 구하는 방법은 알려져 있지 않다. 본 논문에서는 동적 프로그래밍 기법을 적용하여 HMM이나 HMM 네트워크 모형에서 필기를 생성하는 방법을 제시하고, 아울러 HMM의 문제점을 지적한다.