KR950012279A - Probabilistic Stroke Recognition Method by Region Segmentation - Google Patents

Probabilistic Stroke Recognition Method by Region Segmentation Download PDF

Info

Publication number
KR950012279A
KR950012279A KR1019930022390A KR930022390A KR950012279A KR 950012279 A KR950012279 A KR 950012279A KR 1019930022390 A KR1019930022390 A KR 1019930022390A KR 930022390 A KR930022390 A KR 930022390A KR 950012279 A KR950012279 A KR 950012279A
Authority
KR
South Korea
Prior art keywords
stroke
region
strokes
recognition
input
Prior art date
Application number
KR1019930022390A
Other languages
Korean (ko)
Other versions
KR100313993B1 (en
Inventor
조문증
Original Assignee
이헌조
주식회사 금성사
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 이헌조, 주식회사 금성사 filed Critical 이헌조
Priority to KR1019930022390A priority Critical patent/KR100313993B1/en
Publication of KR950012279A publication Critical patent/KR950012279A/en
Application granted granted Critical
Publication of KR100313993B1 publication Critical patent/KR100313993B1/en

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Theoretical Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • Artificial Intelligence (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Character Discrimination (AREA)

Abstract

본 발명은 온라인 필기체 인식의 핵심기술인 획인식 방법에 관한 것으로, 복잡한 굴곡이 있는 획에 대해서는 정확하게 인식할 수 없는 단점이 있다. 왜냐하면, 획의 굴곡이 많아지게 되면 대분류가 부정확하게 될 뿐만 아니라 특정값 자세가 획의 굴곡에 대한 정보를 자세하게 가질 수 없으므로 세밀한 비교분석이 어려워지기 때문이다. 특히 각 필터링과 혹 제거시 많은 정보를 상실하게 되어 아무리 비교를 잘 하더라도 획의 오인식이 발생되는 문제점이 있었는 바, 본 발명은 이를 해결하기 위해 이미 학습되어 있는 획들과 입력된 획의 비유사도를 출력함으로써 인식획 및 후보획을 결과로 받을 수 있고, 또한 각 획의 비유사도를 측정함으로써 글자 인식에 있어서 비록 획이 오인식되었더라도 다음 후보 획으로 글자를 구성해 보아 오인식을 정정할 수 있게 되며, 특히 후보획을 이용하여 인식된 글자 이외의 후보문자를 발생시킬 수 있게 한 것이다.The present invention relates to a stroke recognition method, which is a core technology of on-line handwriting recognition, and has a disadvantage in that it cannot accurately recognize a stroke having a complex curve. This is because when the strokes are more curved, not only the large classification becomes inaccurate, but also because a specific posture cannot have detailed information about the curves of the stroke, it is difficult to make detailed comparative analysis. In particular, there was a problem in that a stroke misrecognition occurred even if a comparison was good, even though a lot of information was lost when filtering or removing the hump. The present invention outputs dissimilarities between strokes that have already been learned and input strokes. Therefore, the recognition stroke and the candidate stroke can be received as a result, and by measuring the dissimilarity of each stroke, even if the stroke is misperceived, it is possible to correct the misrecognition by constructing the character with the next candidate stroke. It is possible to generate candidate characters other than the recognized characters by using strokes.

Description

영역분할에 의한 확률적인 획 인식 방법Probabilistic Stroke Recognition Method by Region Segmentation

본 내용은 요부공개 건이므로 전문내용을 수록하지 않았음Since this is an open matter, no full text was included.

제2도는 본 발명의 영역분할에 의한 확률적인 획 인식 블록도.2 is a probabilistic stroke recognition block diagram by region division of the present invention.

제3도는 입력획을 전처리와 영역분리한 결과를 보인 것으로, (가)는 입력획에 좌표, (나)는 거리 필터링 후의 입력획에 대한 좌표, (다)는 영역 분리후의 입력획에 대한 좌표.Figure 3 shows the results of preprocessing and region separation of input strokes, (a) the coordinates of the input stroke, (b) the coordinates of the input stroke after distance filtering, and (c) the coordinates of the input stroke after region separation. .

제4도는 8방향 코드의 설명도.4 is an explanatory diagram of an eight-way code.

Claims (4)

시간에 따라 순차적으로 입력되는 입력획의 X, Y정보에 대해 중복된 점들과 일정한 거리내에 존재하는 점들을 제거하는 거리 필터링 과정과, 상기 입력된 획을 이루고 있는 부분중에서 특징으로 작용할 수 있는 영역을 구분하는 영역 분리과정과, 상기 분리된 영역에서 특징값을 추출하는 특징 추출과정과, 상기 특징 추출과정까지의 입력처리과정에 의해 형성된 데이타 베이스로 부터 입력획과 가장 유사한 기준값을 인식결과로 출력하는 인식 과정으로 이루어지는 것을 특징으로 하는 영역분할에 의한 확률적인 획 인식 방법.The distance filtering process of removing overlapping points and points existing within a predetermined distance with respect to X and Y information of an input stroke sequentially input according to time, and an area capable of functioning as a feature among the parts forming the input stroke Outputting a reference value most similar to an input stroke from a database formed by a region separation process for separating, a feature extraction process for extracting feature values from the separated region, and an input processing process up to the feature extraction process; Stochastic stroke recognition method by region division, characterized in that the recognition process. 제1항에 있어서, 영역분리과정은 누적된 각이 90°이상이 되면 하나의 부분으로 분리하는 것을 특징으로 하는 영역분할에 의한 확률적인 획 인식 방법.2. The method of claim 1, wherein the area separation process separates the data into one portion when the accumulated angle becomes greater than 90 °. 제1항에 있어서, 특정추출과정은 영역내 획의 크기, 선택된 영역의 전체 획 내에서의 위치, 영역에서 획의 방향에 관한 정보를 특정값으로 간주하는 것을 특징으로 하는 영역분할에 의한 확률적인 획 인식 방법.The method according to claim 1, wherein the specific extraction process regards information about the size of the stroke in the region, the position in the entire stroke of the selected region, and the direction of the stroke in the region as a specific value. How to recognize strokes. 제1항에 있어서, 인식과정에서 데이타베이스는 영역의 갯수에 따라 결정하는 것을 특징으로 하는 영역분할에 의한 확률적인 획 인식 방법.The method of claim 1, wherein in the recognition process, the database is determined according to the number of regions. ※ 참고사항 : 최초출원 내용에 의하여 공개하는 것임.※ Note: The disclosure is based on the initial application.
KR1019930022390A 1993-10-26 1993-10-26 Method for recognizing probable stroke by area partition KR100313993B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
KR1019930022390A KR100313993B1 (en) 1993-10-26 1993-10-26 Method for recognizing probable stroke by area partition

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
KR1019930022390A KR100313993B1 (en) 1993-10-26 1993-10-26 Method for recognizing probable stroke by area partition

Publications (2)

Publication Number Publication Date
KR950012279A true KR950012279A (en) 1995-05-16
KR100313993B1 KR100313993B1 (en) 2002-02-19

Family

ID=37531364

Family Applications (1)

Application Number Title Priority Date Filing Date
KR1019930022390A KR100313993B1 (en) 1993-10-26 1993-10-26 Method for recognizing probable stroke by area partition

Country Status (1)

Country Link
KR (1) KR100313993B1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100414051B1 (en) * 1995-12-20 2004-03-18 엘지전자 주식회사 Method for recognizing stroke of character

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100414051B1 (en) * 1995-12-20 2004-03-18 엘지전자 주식회사 Method for recognizing stroke of character

Also Published As

Publication number Publication date
KR100313993B1 (en) 2002-02-19

Similar Documents

Publication Publication Date Title
US5854855A (en) Method and system using meta-classes and polynomial discriminant functions for handwriting recognition
US5768417A (en) Method and system for velocity-based handwriting recognition
JP5071914B2 (en) Recognition graph
CN100412861C (en) Apparatus and method for searching for digital ink query
US5802205A (en) Method and system for lexical processing
EP0355748A2 (en) A pattern recognition apparatus and method for doing the same
US11132576B2 (en) Text recognition method and apparatus, electronic device, and storage medium
Biswas et al. Writer identification of Bangla handwritings by radon transform projection profile
JP4958236B2 (en) Method and apparatus for recognizing handwritten patterns
US5659633A (en) Character recognition method utilizing compass directions and torsion points as features
KR950012279A (en) Probabilistic Stroke Recognition Method by Region Segmentation
KR940007345B1 (en) On-line recognitin method of hand-written korean character
JP3457094B2 (en) Character recognition device and character recognition method
KR960013818B1 (en) Hangeul combinational recognition apparatus using recognizing multiple stroke
Lee et al. On-line cursive script recognition using an island-driven search technique
Wehbi et al. Numeral characters and capital letters segmentation recognition in mixed handwriting context
KR940001739B1 (en) On-line hand-written korean character recognition method by recognizing stroke
JPS63136286A (en) Online character recognition system
JPS6022793B2 (en) character identification device
KR930012141B1 (en) Recogntion method for on-line writing down character
KR940003828B1 (en) On-line recognizing method of hand-written korean character
KR100414051B1 (en) Method for recognizing stroke of character
KR930014167A (en) Online Cursive English / Number Recognition System
JP2962984B2 (en) Character recognition device
JP2000090203A (en) Method and device for recognizing character

Legal Events

Date Code Title Description
A201 Request for examination
E902 Notification of reason for refusal
E701 Decision to grant or registration of patent right
GRNT Written decision to grant
LAPS Lapse due to unpaid annual fee