JPS6246039B2 - - Google Patents

Info

Publication number
JPS6246039B2
JPS6246039B2 JP55119365A JP11936580A JPS6246039B2 JP S6246039 B2 JPS6246039 B2 JP S6246039B2 JP 55119365 A JP55119365 A JP 55119365A JP 11936580 A JP11936580 A JP 11936580A JP S6246039 B2 JPS6246039 B2 JP S6246039B2
Authority
JP
Japan
Prior art keywords
frame
character
distance
characters
change point
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.)
Expired
Application number
JP55119365A
Other languages
Japanese (ja)
Other versions
JPS5745676A (en
Inventor
Shinichi Shimizu
Masumi Yoshida
Yukikazu Kaburayama
Akira Inoe
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.)
Fujitsu Ltd
Original Assignee
Fujitsu 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 Fujitsu Ltd filed Critical Fujitsu Ltd
Priority to JP55119365A priority Critical patent/JPS5745676A/en
Publication of JPS5745676A publication Critical patent/JPS5745676A/en
Publication of JPS6246039B2 publication Critical patent/JPS6246039B2/ja
Granted legal-status Critical Current

Links

Landscapes

  • Character Input (AREA)

Description

【発明の詳細な説明】 本発明は帳票上に記入された文字列数字、記
号、等を含むより各文字を切り出す文字切り出し
方式に関するものである。
DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a character extraction method for cutting out each character from a character string including numbers, symbols, etc. written on a form.

従来より手書の文字認識においては各文字列よ
り1文字ずつ文字を分離するのが非常に困難であ
り一字ごとに制限枠を設けて記入者に記入させて
いる。しかしながら制限枠を設ける事は帳票を設
計するに当つて非常に不都合であり出来るだけ帳
票を簡略化したい要望が出ている。
Conventionally, in handwritten character recognition, it has been very difficult to separate each character from each character string, so a limit frame has been set up for each character and the person is asked to fill in the space. However, setting a limit frame is extremely inconvenient when designing forms, and there is a desire to simplify forms as much as possible.

一方文字認識装置においても同様に制限枠なし
でも文字の分離を可能とする装置の開発が要求さ
れている。
On the other hand, there is also a demand for the development of a character recognition device that can similarly separate characters without a restriction frame.

従つて本発明では上記問題点を解決し制限枠な
しでも十分文字の分離が可能な文字の分離方式を
提供する事を目的とするものでこの目的は、一定
巾の帯状の枠内に一列に記入された複数の文字よ
り1文字ごとに文字を切り出す文字の切り出し方
式において、該一方の枠より他方の枠に向かつて
垂直方向の各文字までの距離の変化を検出するヒ
ストグラム作成回路と、該ヒストグラム作成回路
により得られた各枠から各文字までの距離の変化
が所定以上変化した事を検出回路と該変化点検出
回路により検出された各枠からの変化点において
少なくとも枠の列方向の距離内にある変化点を1
つの組として検出する組合せ検出回路とを設け、
該変化点の各組を基準に文字の切り出しを行なう
ようにした事を特徴とする文字の切り出し方式に
より達成する事が出来る。
Therefore, it is an object of the present invention to solve the above-mentioned problems and provide a character separation method that can sufficiently separate characters even without a limiting frame. In a character extraction method for cutting out characters one by one from a plurality of written characters, a histogram creation circuit detects a change in distance to each character in a vertical direction from one frame to another; A detection circuit that detects that the change in distance from each frame to each character obtained by the histogram creation circuit has changed by a predetermined value or more; and at least the distance in the column direction of the frame at the change point from each frame detected by the change point detection circuit. 1 change point within
and a combination detection circuit for detecting as one set,
This can be achieved by a character extraction method characterized in that characters are extracted based on each set of change points.

以下本発明を図面を参照しながら説明する。 The present invention will be explained below with reference to the drawings.

第1図は本発明の文字の切出し方式の一実施例
である。
FIG. 1 shows an embodiment of the character cutting method of the present invention.

図において、1はビデオ観測部、2はビデオバ
ツフア、3はヒストグラム作成回路、4は変化点
検出回路、5は変化点メモリバツフア、6は抽出
点決定回路、7組合せ検索回路、8は文字切り出
し位置決定回路、9は文字認識部をそれぞれ示
す。
In the figure, 1 is a video observation unit, 2 is a video buffer, 3 is a histogram creation circuit, 4 is a change point detection circuit, 5 is a change point memory buffer, 6 is an extraction point determination circuit, 7 is a combination search circuit, and 8 is a character extraction position determination circuit. The circuit and 9 indicate a character recognition section, respectively.

また第2図乃至第8図は第1図における各部の
処理状態を示す。
Further, FIGS. 2 to 8 show the processing status of each part in FIG. 1.

帳票上の1定巾の帯状の枠内に一列に記入され
た文字は光学読取り装置等のビデオ観測部1に読
取られ、ビデオバツフア2に格納される。
Characters written in a line within a strip-shaped frame of one fixed width on a form are read by a video observation unit 1 such as an optical reader and stored in a video buffer 2.

従つてビデオバツフア2内には第2図に示すよ
うに帯状の枠A,B間に1列に記入された文字が
格納される。ただし枠については、読取れる状態
でもよいし仮想的に決定されたものでもよい。
Therefore, in the video buffer 2, characters written in a line between the strip-shaped frames A and B are stored as shown in FIG. However, the frame may be in a readable state or may be determined virtually.

このようにビデオバツフア2内に格納されたビ
デオは順次枠A,Bの列方向に読み出されヒスト
グラム作成回路3に入力される。
The video thus stored in the video buffer 2 is sequentially read out in the column direction of frames A and B and input to the histogram creation circuit 3.

ヒストグラム作成回路3においては第3図示す
ように枠Bから枠Aに向かつて垂直方向の文字ま
での距離(文字にぶつからない場合には枠Aまで
の距離)を列方向に順次検出するととも第4図に
示すように枠Aから枠Bに向かつて同様に垂直方
向の文字までの距離を列方向に順次検出し、この
検出情報を変化点検出回路4に順次送る。変化点
検出回路4においては各枠A,Bから文字までの
距離が予じめ決められた所定以上の変化(この距
離は入力される各文字の高さより決定される。)
した点を順次に第3図及び第4図にB1乃至B3
びA1乃至A11として示すように検出し変化点メモ
リバツフア5に変化点のメモリバツフア5に変化
点のメモリアドレスと距離を各枠A,Bからの距
離別に記憶する。
As shown in Figure 3, the histogram creation circuit 3 sequentially detects the distance to a character in the vertical direction from frame B to frame A (or the distance to frame A if the character does not collide) in the column direction. As shown in FIG. 4, the distances to the characters in the vertical direction are similarly sequentially detected in the column direction from frame A to frame B, and this detected information is sequentially sent to the change point detection circuit 4. In the change point detection circuit 4, the distance from each frame A, B to the character changes by a predetermined value or more (this distance is determined from the height of each input character).
The detected points are sequentially detected as shown as B 1 to B 3 and A 1 to A 11 in FIGS. 3 and 4, and the memory address and distance of each change point are stored in the change point memory buffer 5. Stored separately by distance from frames A and B.

この変化点メモリバツフア5に記憶された各枠
の変化点のメモリアドレスと各枠からの距離は抽
出点決定回路6に入力される。
The memory address of the changing point of each frame and the distance from each frame stored in the changing point memory buffer 5 are input to the extraction point determining circuit 6.

抽出点決定回路6においては各枠A,B上の各
変化点が所定の距離内に複数存在する場合(この
距離は予想される文字の幅より決定され少なくと
も予想される1文字の幅より小さい。)には各変
化点の文字までの距離が1番小さいものを残して
他の変化点を削除し。た後の変化点とその文字ま
での距離を組合せ検出回路7に送られる。(例え
ば第5図A4,A5,A6はA4だけが残る。)すなわ
ち第5図に示すように枠A及び枠Bからの変化点
と文字までの距離が入力される。
In the extraction point determination circuit 6, if there are multiple change points on each frame A and B within a predetermined distance (this distance is determined from the expected width of a character and is smaller than the expected width of at least one character) ), leave the one with the smallest distance to the character at each change point and delete the other change points. The change point after the change and the distance to the character are sent to the combination detection circuit 7. (For example, from A 4 , A 5 , and A 6 in FIG. 5, only A 4 remains.) That is, as shown in FIG. 5, the distances from the points of change and the characters from the frames A and B are input.

ただし第5図は説明をわかりやすくするために
仮想的に図示したものので実際には変化点のメモ
リアドレスと文字までの距離が送られる。
However, FIG. 5 is a virtual diagram to make the explanation easier to understand, and in reality, the memory address of the change point and the distance to the character are sent.

組合せ検出回路7においては枠Aからの変化点
と枠Bからの変化点との組合せを検出する。
The combination detection circuit 7 detects a combination of a change point from frame A and a change point from frame B.

すなわち組合せ検出回路においては例えば第5
図に示す枠Aからの変化点に対して枠Bからの変
化点が列方向の所定の距離内に存在する場合にこ
れを1つの組合せとして検出する。
In other words, in the combination detection circuit, for example, the fifth
If the point of change from frame B shown in the figure is within a predetermined distance in the column direction from the point of change from frame A, this is detected as one combination.

この場合にはA1、とB1、A2とB2、A4とB3、A7
とB5、A9とB6、A11とB8である。
In this case A 1 , and B 1 , A 2 and B 2 , A 4 and B 3 , A 7
and B 5 , A 9 and B 6 , and A 11 and B 8 .

このこの様にして組合せを決定した後、組合の
決定された変化点には対応が取れる形でフラグを
付加して抽出点決定回路6内容とともに文字切り
出し位置決定回路8に送る。
After determining the combination in this manner, a flag is added to the determined change point of the combination in a manner that allows correspondence, and the flag is sent to the character cutout position determination circuit 8 together with the contents of the extraction point determination circuit 6.

文字切り出し位置決定回路8においては以下の
如く文字切り出しをビデオバツフア2に対して指
示する。
The character cutting position determination circuit 8 instructs the video buffer 2 to cut out characters as follows.

この文字の切り出しに際しては各枠A,Bから
の変化点と距離及びその関係が以下の4通りの場
合が生じ以下の処理を行なう。
When cutting out this character, there are four cases in which the change points and distances from each frame A and B and their relationships are as follows, and the following processing is performed.

a 組合せ検出回路7において決定された組合せ
の各変化点から文字までの距離を加算した場合
に枠の幅よりも長くなる場合(第6図参照)。
a. When the distance from each change point of the combination determined by the combination detection circuit 7 to the character is added up to be longer than the width of the frame (see FIG. 6).

b aと反対に短かくなる場合(第7図参照)。b If it becomes shorter, contrary to a (see Figure 7).

c 上記a,b、と同様決定された組合せの変化
点が列方向に対して同一位置にある場合(第5
図A1とB1、A4とB3、A11、B8の場合)。
c When the change points of the combinations determined in the same way as a and b above are at the same position in the column direction (fifth
For figures A 1 and B 1 , A 4 and B 3 , A 11 , B 8 ).

d 組合せ検出回路7で組合せがなかつた変化点
(第8図参照)。
d A change point where there is no combination in the combination detection circuit 7 (see FIG. 8).

上記aの場合に第6図1に示すように上の枠A
を基準として下の枠Bまでの距離をY、枠Aから
の変化点から文字までの距離y1、枠Bからの変化
点から文字までの距離をy2とすると中心点ys
は、 ys=y−y+Y/2 となり、また上枠Aの変化点の位置をx1、下枠B
の変化点の位置をx2とすると枠Aからy2sまでは
x1の位置で文字を分離し、ysから枠Bまではx2
の位置で分離するようにビデオバツフア2に対し
てメモリアドレスを指示する。
In the case of above a, as shown in Fig. 6, the upper frame A
If the distance to the lower frame B is Y, the distance from the change point from frame A to the character is y 1 , and the distance from the change point from frame B to the character is y 2 , then the center point y s
is y s =y 1 −y 2 +Y/2, and the position of the change point of upper frame A is x 1 and lower frame B
If the position of the change point is x 2 , then from frame A to y 2s
Separate the characters at x 1 position, x 2 from y s to frame B
A memory address is instructed to the video buffer 2 so as to separate at the position.

従つて第6図2に示すような形で文字を分離し
切り出す事が可能となる。
Therefore, it is possible to separate and cut out characters in the form shown in FIG. 62.

上記bの場合には第7図1に示すように枠Aか
らY−y2までがx1、(Y−y2)から枠Bまでをx2
位置で分離するようにビデオバツフア2にアドレ
スを指示する事により第7図2に示すように文字
を分離し切り出す事が出来る。
In the case of b above, addresses are given to the video buffer 2 so that frames A to Y-y 2 are separated by x 1 , and frames (Y-y 2 ) to frame B are separated by x 2 , as shown in Fig. 7-1. By instructing, characters can be separated and cut out as shown in FIG. 72.

上記cの場合には無条件に各変化点間で文字を
分離するようにビデオバツフア2にアドレスを指
示し文字を分離切り出す。
In case c above, an address is given to the video buffer 2 to unconditionally separate the characters between each change point, and the characters are separated and cut out.

上記dの場合には第8図のように一方の枠Aか
らの変化点から垂直方向に切る事により文字を分
離する。
In the case of d above, the characters are separated by cutting vertically from the point of change from one frame A, as shown in FIG.

このようにして文字の切り出し位置検出回路8
ビデオバツフア2に対して文字の切り出しアドレ
スを指示し、切り出された一文字ごとの文字を文
字認識部9に送り文字認識を行なう。
In this way, the character cutting position detection circuit 8
A character extraction address is instructed to the video buffer 2, and each character extracted is sent to the character recognition section 9 for character recognition.

以上のように本発明は各文字ごとの制限枠を帳
票上に設けなくとも容易に一文字ずつの文字の切
り出しを可能としたため、制限枠を必要としない
帳票に記入者が自由に書く事を可能としたため帳
票設計が簡略化されるとともに文字認識装置にお
いても文字の切り出しが容易になつたため認識率
も向上させる事が出来る。
As described above, the present invention makes it possible to easily cut out characters one by one without setting a limit frame for each character on a form, and thus allows the person filling in the form to write freely on a form that does not require a limit frame. This simplifies the form design and also makes it easier for the character recognition device to cut out characters, thereby improving the recognition rate.

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

第1図は本発明の文字の切り出し方式の一実施
例で第2図乃至第8図は第1図における各部の処
理工程をそれぞれ示し、さらに図において1はビ
デオ観測部、2はビデオバツフア、3はヒストグ
ラム作成回路、4は変化点検出回路、5は変化点
メモリバツフア、6は抽出点決定回路、7は組合
せ検出回路、8は文字切り出し位置決定回路、9
は文字認識部をそれぞれ示す。
FIG. 1 shows an embodiment of the character extraction method of the present invention, and FIGS. 2 to 8 show the processing steps of each part in FIG. is a histogram creation circuit, 4 is a change point detection circuit, 5 is a change point memory buffer, 6 is an extraction point determination circuit, 7 is a combination detection circuit, 8 is a character cutting position determination circuit, 9
indicate character recognition units, respectively.

Claims (1)

【特許請求の範囲】[Claims] 1 帳票上の一定巾の帯状の枠内に一列に記入さ
れた複数の文字を読取り、該読取られた複数の文
字より一文字ごとに文字を切り出す文字の切り出
し方式において、該一方の枠より他方の枠に向か
つ垂直方向の各文字までの距離の変化と該他方の
枠より一方の枠に向かつて垂直方向の各文字まで
の距離の変化を検出するヒストグラム作成回路、
該ヒストグラム作成回路により得られた各枠から
文字までの距離の変化が所定以上変化した事を検
出する変化点検出回路、該変化点検出回路により
該一方の枠側にて検出された第1の変化点と他方
の枠側にて検出された第2の変化点との間の少な
くとも列方向の距離が所定距離内にある一対の変
化点を1つの組として検出する組合せ検出回路と
を備え、上記一対の変化点の位置に対応して文字
の切り出し位置を決定するようにした事を特徴と
する文字の切り出し方式。
1 In a character extraction method in which a plurality of characters written in a line within a band-shaped frame of a certain width on a form are read and characters are cut out one by one from the plurality of read characters, one frame is cut out from the other. a histogram creation circuit that detects changes in the distance to each character in the vertical direction toward the frame and changes in the distance to each character in the vertical direction from the other frame toward one frame;
a change point detection circuit for detecting that the distance from each frame to the character obtained by the histogram creation circuit has changed by more than a predetermined value; a combination detection circuit that detects as one set a pair of change points in which the distance between the change point and the second change point detected on the other frame side is within a predetermined distance at least in the column direction; A character cutting method characterized in that a character cutting position is determined in accordance with the positions of the pair of change points.
JP55119365A 1980-08-29 1980-08-29 Cut-out system of character Granted JPS5745676A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP55119365A JPS5745676A (en) 1980-08-29 1980-08-29 Cut-out system of character

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP55119365A JPS5745676A (en) 1980-08-29 1980-08-29 Cut-out system of character

Publications (2)

Publication Number Publication Date
JPS5745676A JPS5745676A (en) 1982-03-15
JPS6246039B2 true JPS6246039B2 (en) 1987-09-30

Family

ID=14759686

Family Applications (1)

Application Number Title Priority Date Filing Date
JP55119365A Granted JPS5745676A (en) 1980-08-29 1980-08-29 Cut-out system of character

Country Status (1)

Country Link
JP (1) JPS5745676A (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS6070536A (en) * 1983-09-28 1985-04-22 Nippon Columbia Co Ltd Optical information recording medium
JPS60142784A (en) * 1983-12-29 1985-07-27 Fujitsu Ltd Character separating system
JP7074904B1 (en) 2021-03-16 2022-05-24 東海カーボン株式会社 Graphite electrode, electric furnace

Also Published As

Publication number Publication date
JPS5745676A (en) 1982-03-15

Similar Documents

Publication Publication Date Title
JP3302147B2 (en) Document image processing method
CN111652144A (en) Topic segmentation method, device, equipment and medium based on target region fusion
JPS6246039B2 (en)
JPS6126150A (en) Registering and retrieving device of document picture file
JPH0423185A (en) Table reader provided with automatic cell attribution deciding function
JPS6336037B2 (en)
JP2618468B2 (en) Document processing device
JPH06131497A (en) Table recognition system
JP3379663B2 (en) Character recognition device
JPS6331830B2 (en)
JPH02305116A (en) Method of information data compression
JPS62243067A (en) Image file device
JP3668026B2 (en) Publication electronic processing equipment
JP2901407B2 (en) Mask resistance direction recognition method
JPS5866148A (en) Discriminating system for opening and closing of rule
JPS596419B2 (en) Character extraction method
JPH01130293A (en) Document image analyzing system
JPH083827B2 (en) Character image processing method
JPH07104940B2 (en) Figure recognition device
JPS60254381A (en) Extracting method of character area
JPS61206090A (en) Character reading device
JPH03161888A (en) Optical character reader
JPS6236274B2 (en)
JPS61263365A (en) Extraction system for character zone using hierarchical processing
JPS589292A (en) Reading system for read only memory device