JPS5835674A - Extracting method for feature of online hand-written character - Google Patents

Extracting method for feature of online hand-written character

Info

Publication number
JPS5835674A
JPS5835674A JP56132633A JP13263381A JPS5835674A JP S5835674 A JPS5835674 A JP S5835674A JP 56132633 A JP56132633 A JP 56132633A JP 13263381 A JP13263381 A JP 13263381A JP S5835674 A JPS5835674 A JP S5835674A
Authority
JP
Japan
Prior art keywords
stroke
memory
register
strokes
polygonal
Prior art date
Legal status (The legal status 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 status listed.)
Granted
Application number
JP56132633A
Other languages
Japanese (ja)
Other versions
JPS638513B2 (en
Inventor
Yasuo Ichinoto
市野渡 康夫
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Oki Electric Industry Co Ltd
Original Assignee
Oki Electric Industry Co Ltd
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 Oki Electric Industry Co Ltd filed Critical Oki Electric Industry Co Ltd
Priority to JP56132633A priority Critical patent/JPS5835674A/en
Publication of JPS5835674A publication Critical patent/JPS5835674A/en
Publication of JPS638513B2 publication Critical patent/JPS638513B2/ja
Granted legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/20Image preprocessing

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Character Discrimination (AREA)

Abstract

PURPOSE:To recognize hand-written characters accurately, by calculating the sum of the shortest distances from the start point, the end point, and bend points of the polygonal line approximate pattern of a stroke to other strokes to detect a duplicate write stroke and extracting features for the duplicate write stroke. CONSTITUTION:One-character components of coordinate data are stored in a coordinate memory 1, and a flag bit is set in the last coordinates of strokes. Strokes of the memory 1 are approximated to polygonal lines by a polygonal approximating equipment 2, and coordinate values of their start points, end points, and bend points of polygonal lines are stored in a polygonal line register memory 3. A value of the shortest distance between coordinate values in the memory 1 and the first coordinates in the memory 3 is detected by a distance calculator 4 and is applied to an adder 5 and is added to contents of an addition value register 6, and the result is stored in the register 6. Contents of the register 6 are read out into a comparison discriminator 7; and when they are smaller than a certain value, the second stroke is discriminated as the duplicate write stroke of the first stroke, and thus, hand-written characters are recognized accurately.

Description

【発明の詳細な説明】 本発明はタブレット上で描かれる手書文字を認識する事
を目的とする手書文字特徴抽出方法に関するものである
DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a handwritten character feature extraction method for the purpose of recognizing handwritten characters drawn on a tablet.

従来この種の特徴抽出においては、文字のストローク数
や各ストロークに関する様々な情報が抽出され文字識別
に利用されるがボールペン等の筆記具によっては文字の
かすれが起りやすく、筆記者がそのストローク上の一部
をもう一度なぞって書く場合がある。この様なときには
ストロークの数や順序が正常に書かれたときと違ってし
まい特徴抽出が正しく行われなくなるという欠点があっ
た。本発明の目的は二重書きを検出しその情報を用いる
事によって上記の欠点を解決するものであり以下詳細に
説明する。
Conventionally, in this type of feature extraction, the number of strokes of a character and various information about each stroke are extracted and used for character identification. You may have to trace some parts over again. In such a case, the number and order of strokes would be different from when they were normally written, resulting in a disadvantage that feature extraction would not be performed correctly. An object of the present invention is to detect double writing and use the information to solve the above-mentioned drawbacks, and will be described in detail below.

添付図面は本発明の実施例のブロック図であってlは座
標メモリ、2は折れ線近似器、3は折れ線登録メモリ、
4は距離計算器、5は加算器、6は加算値レジスタ、7
は比較判定器である。
The attached drawing is a block diagram of an embodiment of the present invention, in which 1 is a coordinate memory, 2 is a polygonal line approximator, 3 is a polygonal line registration memory,
4 is a distance calculator, 5 is an adder, 6 is an addition value register, 7
is a comparison judge.

1文字分の座標データは、サンプル順に座標メモリ1に
格納されていて、ストロークの最後の座標にはストロー
クエンドを示すフラグビットがあり1“がセットされて
いる。折れ線近似器2は各ストロークを折れ線近似化し
その始点、終点、及び折れ線の分割点の座標値を折れ線
登録メモリ3に格納する。ストロークの折れ線近似化に
ついては広く一般的に知られている手法であるのでここ
では説明を省略する。
The coordinate data for one character is stored in the coordinate memory 1 in sample order, and the last coordinate of the stroke has a flag bit indicating the end of the stroke, which is set to 1.The polygonal line approximator 2 The polygonal line is approximated and the coordinate values of the starting point, end point, and dividing point of the polygonal line are stored in the polygonal line registration memory 3. Since the polygonal line approximation of the stroke is a widely known method, the explanation will be omitted here. .

次に距離計算器4は第1ストロークに対応する座標メモ
リ1内の各座標値と第2ストロークに関する折れ線登録
メモリ3の第1番目の座標値との距離の最も小さいもの
を検出しその値を加算器5によって加算値レジスタ6の
内容と加算し、結果を加算値レジスタ6に格納する。加
碧値レジスタ6の内容は最初にクリアされているので第
1回目の検出値はそのまま加算値レジスタ6に格納され
る。第2番目の座標値につ℃・ても同様に第1ストロー
クの各座標値との最小値を検出し加算値レジスタ6に加
算する。このように第2ストロークに対応する折れ線登
録メモリ3の内容に関しての処理が終了したら比較判定
器7は′加算値レジスタ6の内容を読み出しこれを一定
の値と比較し小さい場合は第2ストロークは第1ストロ
ークの2重書きストロークと判定し座標メモリl内の第
2ストロークのデータに2重書きストロークである事を
示すフラグビットに11“を書き込む。またそうでない
場合は知“を書き込む。引き続き第2ストロークに対応
する座標メモリ1の内容と第3ストロークに対応する折
れ線登録メモリ3の内容についても同様の処理を行い結
果によって第3ストロークの2屯書きストロークを示す
フラグビットに11″または0“を書き込む。このよう
にして処理を繰り返し文字の最終のひとつ前のストロー
クと最終ストロークについての処理まで終了すると座標
メモリ1の上には前に描いたストロークの上をもう一度
なそって描いたストロークの座標データにはフラグビッ
トによって二重書きストロークであることが示されてい
る状態になる。
Next, the distance calculator 4 detects the smallest distance between each coordinate value in the coordinate memory 1 corresponding to the first stroke and the first coordinate value in the polygonal line registration memory 3 regarding the second stroke, and calculates that value. The adder 5 adds the contents of the added value register 6 and stores the result in the added value register 6. Since the contents of the added value register 6 are initially cleared, the first detected value is stored as is in the added value register 6. Similarly, for the second coordinate value, the minimum value with respect to each coordinate value of the first stroke is detected and added to the addition value register 6. When the processing regarding the contents of the polygonal line registration memory 3 corresponding to the second stroke is completed in this way, the comparator 7 reads the contents of the added value register 6 and compares it with a certain value, and if it is smaller, the second stroke is It is determined that the first stroke is a double writing stroke, and 11" is written in the flag bit indicating that it is a double writing stroke in the data of the second stroke in the coordinate memory l. If it is not, " is written. Subsequently, the same process is performed for the contents of the coordinate memory 1 corresponding to the second stroke and the contents of the polygonal line registration memory 3 corresponding to the third stroke, and depending on the result, the flag bit indicating the second stroke of the third stroke is set to 11'' or Write 0". When the process is repeated in this way and the process is completed for the stroke before the last stroke of the character and the final stroke, the coordinate data of the stroke drawn by tracing over the previously drawn stroke again is stored in the coordinate memory 1. is in a state where the flag bit indicates that it is a double writing stroke.

リド説明したように、本実施例では筆記者が一度描(ゴ
ニストローク上の一部をもう一度なぞって描・、・た場
合でもその情報が座標データ上に残される為特徴抽出の
際にそのストロークに関する情報?無視するなどの手法
を用いる事により正常な特徴抽出が行えるという利点が
ある。
As explained above, in this embodiment, even if the scribe draws once (tracing a part of the goni stroke again, etc.), that information is left on the coordinate data, so when extracting features, the stroke There is an advantage that normal feature extraction can be performed by using techniques such as ignoring information about information.

本発明はストロークの折れ線近似バタンにおけろ始7壱
、終点、分割点と他のストロークとの最短距離の和を計
算することによって二重書きストロークを検出している
ので文字のストロークの中に二重に書かれたストローク
があった場合でも正常な特徴抽出が行えるという利点が
あるので数字、英字、カナその他の記号などのオンライ
ン手書文字認識に利用することができる。
The present invention detects double writing strokes by calculating the sum of the shortest distances between the starting point, ending point, dividing point, and other strokes in the stroke polygonal approximation button. This method has the advantage of correctly extracting features even when there are double strokes, so it can be used for online handwritten character recognition such as numbers, alphabets, kana, and other symbols.

【図面の簡単な説明】[Brief explanation of the drawing]

添付図面は本発明の実施例のブロック圀である。 1・・・座標メモリ、 2・・・折れ線近似器、3・・
・折れ線登録メモリ、4・・・距離計算器、5・・・加
 算 器、  6・・−加算値レジスタ、7・・・比較
判定器。 特許出願人 沖電気工業株式会社 特許出願代理人 弁理士 山 本 恵 − !−1釉4・A−1,awh鼾
The accompanying drawings are block diagrams of embodiments of the invention. 1... Coordinate memory, 2... Polygonal line approximator, 3...
- Line registration memory, 4... distance calculator, 5... adder, 6... - addition value register, 7... comparison judger. Patent applicant Oki Electric Industry Co., Ltd. Patent application agent Megumi Yamamoto -! -1 glaze 4・A-1,awh snoring

Claims (1)

【特許請求の範囲】[Claims] 1文字分の座標データをストロークに分割し、各ストロ
ークの折れ線近似化パタンの始点、終端及び折れ線分側
点と、別のストロークとの距離が所定の値より小さいど
き、折れ線近似化した方のストロークが別のストローク
に対する二重書きストロークであるとして、文字識別に
利用することを特徴とするオンライン手書文字特徴抽出
方法。
When the coordinate data for one character is divided into strokes, and the distance between the start point, end, and side point of the polyline segment of each stroke's polygonal line approximation pattern and another stroke is smaller than a predetermined value, the polyline approximation pattern is An online handwritten character feature extraction method characterized in that the stroke is used for character identification, assuming that the stroke is a double writing stroke with respect to another stroke.
JP56132633A 1981-08-26 1981-08-26 Extracting method for feature of online hand-written character Granted JPS5835674A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP56132633A JPS5835674A (en) 1981-08-26 1981-08-26 Extracting method for feature of online hand-written character

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP56132633A JPS5835674A (en) 1981-08-26 1981-08-26 Extracting method for feature of online hand-written character

Publications (2)

Publication Number Publication Date
JPS5835674A true JPS5835674A (en) 1983-03-02
JPS638513B2 JPS638513B2 (en) 1988-02-23

Family

ID=15085885

Family Applications (1)

Application Number Title Priority Date Filing Date
JP56132633A Granted JPS5835674A (en) 1981-08-26 1981-08-26 Extracting method for feature of online hand-written character

Country Status (1)

Country Link
JP (1) JPS5835674A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4607386A (en) * 1983-10-06 1986-08-19 Sharp Kabushiki Kaisha Handwritten character recognition device
JPS61226886A (en) * 1985-03-30 1986-10-08 Toshiba Corp Pattern input device
JPH03127291A (en) * 1989-10-13 1991-05-30 Internatl Business Mach Corp <Ibm> Apparatus and method for recognizing hand writing stroke
JP2013050785A (en) * 2011-08-30 2013-03-14 Fuji Xerox Co Ltd Information processing device, imaging device, and program

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4607386A (en) * 1983-10-06 1986-08-19 Sharp Kabushiki Kaisha Handwritten character recognition device
JPS61226886A (en) * 1985-03-30 1986-10-08 Toshiba Corp Pattern input device
JPH03127291A (en) * 1989-10-13 1991-05-30 Internatl Business Mach Corp <Ibm> Apparatus and method for recognizing hand writing stroke
JP2013050785A (en) * 2011-08-30 2013-03-14 Fuji Xerox Co Ltd Information processing device, imaging device, and program

Also Published As

Publication number Publication date
JPS638513B2 (en) 1988-02-23

Similar Documents

Publication Publication Date Title
US4542526A (en) Character recognition apparatus
JPS6079485A (en) Handwriting character recognition processing device
CN111126160B (en) Intelligent Chinese character structure evaluation method and system constructed based on five-stroke input method
JPS5835674A (en) Extracting method for feature of online hand-written character
JPH0520794B2 (en)
JP2671984B2 (en) Information recognition device
JPS62154078A (en) Extraction system for feature point in recognition of online hand-written character
JPS5840236B2 (en) Online handwritten character recognition method
JPS6022793B2 (en) character identification device
KR900005141B1 (en) Handwritter character recognizing device
JP2519782B2 (en) Character separation method
JPH01213772A (en) On-line character recognizing system
JP2851865B2 (en) Character recognition device
JPS62162182A (en) Extraction system for bending point of hand-written picture input device
JPH0438024B2 (en)
JPS63301383A (en) Handwritten character recognition device
JPS62160597A (en) On-line consecutive character recognition device
JPS6113386A (en) Recognizing device of hand-written character
JPS6327753B2 (en)
JPS6186881A (en) Recording system for on-line handwritten character
JPH0443316B2 (en)
JPH063613B2 (en) Handwriting recognition device
JPS6079482A (en) On-line recognition processing system
JPS5668879A (en) Real-time hand-written character recognition system
JPH01231189A (en) Character recognizing system