JPS5856079A - Character segmenting device of optical character reader - Google Patents

Character segmenting device of optical character reader

Info

Publication number
JPS5856079A
JPS5856079A JP56154089A JP15408981A JPS5856079A JP S5856079 A JPS5856079 A JP S5856079A JP 56154089 A JP56154089 A JP 56154089A JP 15408981 A JP15408981 A JP 15408981A JP S5856079 A JPS5856079 A JP S5856079A
Authority
JP
Japan
Prior art keywords
character
projection
register
data
memory
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.)
Pending
Application number
JP56154089A
Other languages
Japanese (ja)
Inventor
Mamoru Maeda
護 前田
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.)
Ricoh Co Ltd
Original Assignee
Ricoh 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 Ricoh Co Ltd filed Critical Ricoh Co Ltd
Priority to JP56154089A priority Critical patent/JPS5856079A/en
Publication of JPS5856079A publication Critical patent/JPS5856079A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10Character recognition
    • G06V30/14Image acquisition
    • G06V30/148Segmentation of character regions
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10Character recognition

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Character Input (AREA)

Abstract

PURPOSE:To segment a character accurately through a small-sized device by providing a register control part which corrects the difference between a counter value, obtained every time the numbers of scanning lines are changed, and the address value of a projection register, and then deciding on character segmenting lines on the basis of projection data values corrected by the control part. CONSTITUTION:Character data Pd from a reading scanner are stored successively in a pattern memory 1, and then read out by command signals lc and dc for scanning the scanning point of a memory control point laterally and longitudinally. The storage data md read out of the memory 1 are supplied to an arithmetic part 5, and the signals lc and dc are applied to the address correction part 8 and addition part 9 of a register control part 7. The difference between the counted value of the control part 2 and the address value of a projection register 6 is corrected through the control part 7 to apply an address correction signal ra to the register 6, storage data rd from which are applied to a character segmenting decision part 10 to decide on character segmenting lines there, thus outputting a cutting position signal cp between characters.

Description

【発明の詳細な説明】 本発明は、光学文字読取装置(以下、0OR)において
、読取った文字画像を1文字ずつ切出す文字読出装置に
関する。
DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a character reading device for cutting out read character images character by character in an optical character reading device (hereinafter referred to as OOR).

00Rにおいては、読取りた文字画像の認識処理の過程
において文字画像のパターンを射影して文字を7文字と
とに切出すという処理が行われる。
In 00R, in the process of recognizing a read character image, a process of projecting the pattern of the character image and cutting out the characters into seven characters is performed.

文字パターンを射影する方法としては1文字パターン列
に対して垂直方向に射影するという方法が従来では一般
的であった。ところが、W文字画像が手書きである場合
、あるいは横書で印刷したものであっても正確に読取れ
ない場合がある。つまシ1手書文字の場合は隣接する文
字枠を越えて一方の文字がIIシの枠内I/ctで及ぶ
仁とがあり。
Conventionally, a method of projecting character patterns in a direction perpendicular to one character pattern string has been a common method. However, even if the W character image is handwritten or printed horizontally, it may not be able to be read accurately. In the case of handwritten characters with one tsumashi, there is a case where one character extends beyond the adjacent character frame by I/ct within the frame of the second character.

これを垂直方向に射影した場合には文字パターンがあた
か4連続しているが如く認識してしまうととになるから
である。一方、印刷文字の場合は特にイタリ、り体の英
文字などのように字体その本のが一定の傾斜角をもって
配列されている場合に同様々ことが生じる。
This is because if this is projected in the vertical direction, the character pattern will be recognized as if it were four consecutive characters. On the other hand, a similar problem occurs in the case of printed characters, especially when the fonts are arranged at a certain angle of inclination, such as in italized or cursive English characters.

例えば、@/図に示すようにイタリック体の(字を垂直
方向に射影して切出し線すを設定した場合には % z
 Iの上部と%H#の下部が重かり、結局。
For example, if you set the cutting line by vertically projecting the italic character (@/ as shown in the figure), % z
The top of I and the bottom of %H# are heavy, so in the end.

連続した文字パターンであると認識されるおそれがある
There is a risk that it will be recognized as a continuous character pattern.

かかる不都合を解決する手段として、斜め方向に射影す
るととKよシ切出線を傾けて誤5iIi識を防止する文
字切出方式がある。しかし、従来の切出し方式の場合は
その構成上シ7トレジス!を多く必髪とするなどに起因
して装置の大型化が余儀なくされるものであった。
As a means to solve this problem, there is a character cutting method in which the cutting line is tilted by K when projected in an oblique direction to prevent the erroneous 5iIi identification. However, in the case of the conventional cutting method, there are 7 registers due to its structure. Due to the large amount of hair required, the device had to be made larger.

〈発明の目的〉 そこで1本発明は、装置構成を大型化するとと々く、正
確に文字を切出すことができる文字切出し装置を提供す
ることを目的とする。
<Objective of the Invention> Accordingly, an object of the present invention is to provide a character cutting device that can accurately cut out characters even when the device configuration is increased in size.

〈発明の構成〉 上記目的を達成するためK、本発明の文字切出装置は。<Structure of the invention> In order to achieve the above object, the present invention provides a character cutting device.

光学文字読取装置において、読取った文字1III儂を
記憶する文字画像パターンメモリと、この文字画像パタ
ーンメモリをアクセスするメモリ制御部ト1文字1ii
fIj!パターンをそのパターン列に対して斜め方向に
射影したときのデータを配憶する射影レジスタと、前記
文字画倫パターンのアクセス時において走査ライン数が
変わるごとにデータカウント値と格納すべき射影レジス
タのアドレス値との差を補正する信号を前記射影レジス
タに出力する射影レジスタ制御部と、射影レジスタから
の射影データに基づいて文字の切出し線を法定する文字
切出し判定部とを備えたことを特徴とするものである。
In an optical character reading device, a character image pattern memory that stores read characters 1III and a memory control unit that accesses this character image pattern memory are provided.
fIj! A projection register that stores data obtained when a pattern is projected diagonally with respect to the pattern sequence, and a projection register that stores data count values every time the number of scanning lines changes when accessing the character/art pattern. The present invention is characterized by comprising a projection register control unit that outputs a signal for correcting a difference with an address value to the projection register, and a character cutting determination unit that determines a character cutting line based on projection data from the projection register. It is something to do.

〈発明の実施例〉 以下1本発明を図示する実施例に基づいて詳述する。第
2図に本発明による文字切出し装置の一実施例をプロ、
り図で示す。
<Embodiments of the Invention> The present invention will be described in detail below based on an illustrative embodiment. FIG. 2 shows an embodiment of the character cutting device according to the present invention.
This is shown in the diagram.

第2図において、符号lは文字画像パターンメモリ(以
下、パターンメモリと略記する。)を示している。この
パターンメモリlには読取スキャナ(図示せず、)の走
査によシ読取られた文字データア。が順次格納される。
In FIG. 2, the symbol l indicates a character image pattern memory (hereinafter abbreviated as pattern memory). This pattern memory 1 stores character data read by scanning with a reading scanner (not shown). are stored sequentially.

第3図はこの文字パターンメモリ/内に文字データ1H
#が格納され九場合の文字データ位置と後述する射影レ
ジスタ1との文応関係、ならびに読出走査に用いられる
走査ポインタPを示したものである。このパターンメモ
リlはメモリ制御部コによりアクセスされ。
Figure 3 shows character data 1H in this character pattern memory.
It shows the correspondence relationship between the character data position and the projection register 1, which will be described later, when # is stored, and the scanning pointer P used for read scanning. This pattern memory l is accessed by the memory control unit.

順次格納データ(文字データ%H’ ) m 、1を出
力する。
Output sequentially stored data (character data %H') m, 1.

メモリ制御部コは、パターンメモIJ /内に設定され
た走査ポインタPC第3図)の位置を横方向に走査する
指命信号d0を出力するデータカウンタ3と、走査ポイ
ンタPの位置を縦方向に走査する信号1゜を出力するラ
インカウンタ弘とを備えて構成される。このメモリ制御
部2によって走査ポインタPは文字パターンメモリl内
を第3図における左から右へ/走査ラインずつ上から下
へと順次走査される。このアクセス動作によりてパター
ンメモリ/から格納データm4が順次読出され。
The memory control unit includes a data counter 3 which outputs an instruction signal d0 to horizontally scan the position of a scanning pointer PC (FIG. 3) set in the pattern memo IJ, and a data counter 3 which outputs an instruction signal d0 to horizontally scan the position of the scanning pointer The line counter outputs a 1° scanning signal. The memory control unit 2 causes the scanning pointer P to sequentially scan the character pattern memory 1 from left to right in FIG. 3/scanning line by scanning line from top to bottom. By this access operation, the stored data m4 is sequentially read out from the pattern memory.

演算部夕に送られること1Icfkる。一方、横方向走
査指令信号d0と縦方向指令信号1゜は後述するレジス
タ制御部7へも送られる。
The data is sent to the arithmetic unit 1Icfk. On the other hand, the horizontal direction scanning command signal d0 and the vertical direction command signal 1° are also sent to the register control section 7, which will be described later.

レジスタ制御部7は後述する射影レジスタtKおける射
影データのアドレスを制御するものであし、アドレス補
正部lと、この補正部tがらの補正信号Cとデータカウ
ンタ参からの縦方向走査指令信号d0とを加算する加算
部りな備えて構成される。加算部りから出方されるアド
レス制御部信号raは、&直方向の射影値を求める場合
と斜め方向の射影値を求める場合とではその内容は異な
る。すなわち。
The register control section 7 controls the address of the projection data in the projection register tK, which will be described later, and uses an address correction section 1, a correction signal C from the correction section t, and a vertical scanning command signal d0 from the data counter. It is composed of an adding section that adds up. The address control unit signal ra output from the adder has different contents depending on whether a projected value in the &orthogonal direction is determined or when a projected value in an oblique direction is determined. Namely.

の 垂直方向の射影値を求める場合には、データカウン
タ参のカウント値すなわち縦方向走査指令信号d0をそ
のままアドレス制御信号として射影レジスタ制御部る。
When calculating the vertical projection value, the count value of the data counter, that is, the vertical scanning command signal d0, is directly used as an address control signal by the projection register control section.

■ 斜め方向の射影値を求める場合には、第1図かられ
かるように、走査ライン数が増加するととにデータカウ
ント値(体。)のアドレス値がその斜角によシずれてく
るため、演算部!がらの演算データPaを走査ライン数
の増加、すなわちラインカウント値(1o)の増加に伴
たってデータカウント値(do)のアドレス値と格納す
べき射影レジスタ基のアドレス値との差を補正する必要
がある。そこで、第4図に示すように1例えば文字パタ
ーン列を水平方向とすると、この水平方向に対して7j
’の射影データを求めるためには。
■ When calculating the projection value in the diagonal direction, as shown in Figure 1, as the number of scanning lines increases, the address value of the data count value (field) shifts according to the diagonal angle. , arithmetic section! It is necessary to correct the difference between the address value of the data count value (do) and the address value of the projection register base to be stored as the number of scan lines increases, that is, the line count value (1o) increases. There is. Therefore, as shown in FIG. 4, if the character pattern string is oriented horizontally, then
To obtain the projection data of '.

tan 71’二参 と近似して、走査ライン数であるラインカウント値(1
゜)がダライン増すととにデータカウント値(do)の
アドレス値アをlずりずらしていけばよい、第参図にお
いて、P−j〜F + 、2はデータカウント値のアド
レス値の例示であり、L〜L+yはラインカウント値の
例示である。この@によれば。
Approximately, the line count value (1
As ゜) increases by a line, the address value A of the data count value (do) can be shifted by l. Yes, and L to L+y are examples of line count values. According to this @.

ライン番号” −”+/ ’ ”+J ’ Xk+jと
インクリメントしてきたと@ifc、データカウント値
P + tにおける演算データ?、の内容が射影レジス
タ基の1’+、@mK格納されることkなる1次に、ラ
イ:/番号”+1.L+よ、L+乙、L+7とインクリ
メントしてきたときに、データカウント値P&Cおける
演算テークPaの内容が射影レジスタ1のP + を番
地に格納されることとなる。このような補正は、レジス
タ制御部7の補正部fKよって行なわれる。補正部fv
ROMにより構成し。
When the line number "-"+/'"+J" is incremented as Next, lie:/number”+1. When incrementing is performed in the order of L+, L+B, and L+7, the contents of the calculation take Pa in the data count value P&C will be stored at the address P + of the projection register 1. Such correction is performed by the correction section fK of the register control section 7. Correction section fv
Constructed by ROM.

入力されるラインカウント値(1゜)K応じた補正値C
0を予め格納してお社ばよい、上述の走査ライン数ダに
対して−lの補正を行う場合の−正値として@j図に示
すような値を格納してかけばよい、ただし、これは斜め
方向忙射影する場合の館であり、fi直力方向射影の場
合は補正値はゼロで&也乙障証懐C0が加算部りにおい
てデータカウント値d 忙加算され、アドレス制II信
号r、とじて射影レジスタ6に与えられるのである。
Correction value C according to the input line count value (1°) K
You can store 0 in advance and multiply it by storing a value as shown in the diagram @j as a -positive value when performing -l correction for the number of scanning lines mentioned above.However, This is the case in the case of diagonal projection, and in the case of direct projection, the correction value is zero and the data count value d is added in the addition section, and the address system II signal r, and is given to the projection register 6.

表お、ここでの説明は切出11aを直線として説明して
いるが、補正値0゜は各ラインととに任意の値を設定で
きるため、ti直方向、斜め方向だけで橙〈、任意の曲
線に対しても追従させること亀できる。その−例を第4
図に示す。
In this table, the cutout 11a is explained as a straight line, but since the correction value 0° can be set to any value for each line, it is possible to set the correction value 0° to any value for each line, so it is possible to It is also possible to follow the curve. The fourth example is
As shown in the figure.

演算部!は、パターンメモリーからの格納デー1−を入
力として射影データを求めるために入力格納データm4
の個々について白画素か黒画素かの判定を行う、そして
、射影のヒストグラムを求める場合には射影レジスタの
出力データr4と格納データno和を求め、その演算デ
ータP6を射影レジスタ4に再入力する。また、射影方
向に黒画素があるかどうかを求める場合には両者の論理
和を算出する。ここで、演算部!の動作の一例を説明す
る。第7図に示すようにメモリー内を順次走査してその
t!−直射形データを集めるとする。
Arithmetic section! input stored data m4 to obtain projection data using stored data 1- from the pattern memory as input.
It is determined whether each pixel is a white pixel or a black pixel, and when obtaining a projection histogram, the output data r4 of the projection register and the stored data no are summed, and the calculated data P6 is re-inputted into the projection register 4. . Furthermore, when determining whether there is a black pixel in the projection direction, the logical sum of the two is calculated. Here, the calculation part! An example of the operation will be explained. As shown in FIG. 7, the memory is sequentially scanned and the t! - Suppose we want to collect direct data.

メモリアドレスは、(列番号1行番号)として表わす、
走査ポインターPの位置が(J、/)番地の・で示す番
地にあるとすると、1行目C行番号)の走査は終了して
いるから、射影レジスタの内容は、 00/10/と表
りている。(J、/)番地の内容はl(黒)であるから
射影レジスタ6の内容を+7する。このため、射影レジ
スタto3番地の内容を演算部の纂l入力r1とし、メ
モリMの(J、/)11地の内容を第コ入力1114と
し1両人力の和Paを出力する。和−は、射影レジスタ
基の3番地へ格納され、内容はコとなる。その後。
The memory address is expressed as (column number 1 row number),
Assuming that the position of the scanning pointer P is at the address (J, /) indicated by *, scanning of the first line (C line number) has been completed, so the contents of the projection register are expressed as 00/10/. There is Since the content of address (J, /) is l (black), the content of projection register 6 is incremented by +7. Therefore, the contents of the projection register to3 are set as the input r1 of the calculation section, and the contents of the (J, /)11 location of the memory M are set as the co-input 1114, and the sum Pa of both hands is output. The sum - is stored at address 3 of the projection register base, and the content becomes ko. after that.

走査ポインターPは(ダ、/)番地へ移動し、同様のこ
とが打力われる。このようKしてメモリl内の全走査が
終了すると、射影レジスタJKは0024110の射影
値が記録される。
The scanning pointer P moves to the (da, /) address and the same thing is input. When all the scans in the memory I are completed in this manner, a projection value of 0024110 is recorded in the projection register JK.

このようkして、射影データが射影レジスタtに格納さ
れたのち、その格納データr4は文字切出し判定部10
K送られる6文字切出し判定部lOは格納データr6の
内容から切出し位置を判定し。
After the projection data is stored in the projection register t in this way, the stored data r4 is transferred to the character segmentation determination unit 10.
K is sent to the 6-character cutout determining unit IO, which determines the cutout position from the contents of the stored data r6.

隣接文字相互間の切断位置信号Opを出力する。A cutting position signal Op between adjacent characters is output.

〈発明の効果〉 以上の通勤1本発明によれば、射影データの数に応じた
シフトレジスタを設ける必要はなく、比較的簡単なディ
ジタル電子回路によって構成することができる。よりて
文字の連結等のWAf!識をしやすい原文字画曹を斜め
方向に射影することによって正111KN識するための
文字切出し装置を小型で構成することができる。
<Effects of the Invention> According to the present invention, there is no need to provide shift registers corresponding to the number of projection data, and the system can be constructed using a relatively simple digital electronic circuit. WAf for connecting characters, etc. By projecting the easy-to-recognize original character stroke in an oblique direction, it is possible to construct a small-sized character cutting device for recognizing normal 111KN.

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

91図はイタリック体の英文字の態様と文字切出し線と
の関係を示す説明図、第2図は本発明による文字切出し
装皺の一実施例を示すプロ、り図。 第3図はパターンメモリと射影レジスタとの関係を示す
観明図、第μ図は斜め方向の射影時における格納アドレ
スの補正状態を示す説明図、第1図は補正部に格納され
た軸正値の例を示す説明図。 第4図は他の切出線を示す説明図、第7図は演算部の動
作例を示す説明図である。 l・・・文字画像パターンメモリ、−・・・メモリ制御
部、6・・・射影レジスタ、7・・・レジスタ制御部。 10・・・文字切出し判定部。 出願人代理人  猪  股     清馬2図 、    帛3図 苓4図       5$5図 )::: L+4− 帛7図 馬6図
FIG. 91 is an explanatory diagram showing the relationship between the aspect of italicized English letters and character cutting lines, and FIG. 2 is a professional drawing showing one embodiment of the character cutting and folding arrangement according to the present invention. Fig. 3 is a visual diagram showing the relationship between the pattern memory and the projection register, Fig. μ is an explanatory diagram showing the correction state of storage addresses when projecting in an oblique direction, and Fig. An explanatory diagram showing an example of values. FIG. 4 is an explanatory diagram showing another cutting line, and FIG. 7 is an explanatory diagram showing an example of the operation of the calculation section. l...Character image pattern memory, -...Memory control unit, 6...Projection register, 7...Register control unit. 10...Character extraction determination section. Applicant's agent Kiyoma Inomata (Figure 2, Figure 3, Figure 4, Figure 5, $5) ::: L+4- Figure 7, Figure 6

Claims (1)

【特許請求の範囲】[Claims] 光学文字断取装置において、読取った文字画像を記憶す
る文字画像パターンメモリと、この文字画像パターンメ
モリをアクセスするメモリ制御部と1文字画像パターン
をそのパターン列に対して斜め方向に射影したときのデ
ータを配憶する射影レジスタと、前記文字画像パターン
のアクセス時において走査ライン数が変わるごとにデー
タカウンタ値と格納すべき射影レジスタのブトレス値と
の差を補正する信号を前記射影レジスタに出力する射影
レジスタ制御部と、射影レジスタからの射影データに基
づいて文字の切出し線を決定する文字切出し判定部とを
備えたことを特徴とする文字切出し装置。
In an optical character cutting device, there is a character image pattern memory that stores read character images, a memory control unit that accesses this character image pattern memory, and a character image pattern when a single character image pattern is projected diagonally with respect to the pattern row. A projection register that stores data, and a signal that corrects a difference between a data counter value and a buttress value of the projection register to be stored each time the number of scanning lines changes when accessing the character image pattern is output to the projection register. A character cutting device comprising: a projection register control section; and a character cutting determination section that determines a character cutting line based on projection data from the projection register.
JP56154089A 1981-09-29 1981-09-29 Character segmenting device of optical character reader Pending JPS5856079A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP56154089A JPS5856079A (en) 1981-09-29 1981-09-29 Character segmenting device of optical character reader

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP56154089A JPS5856079A (en) 1981-09-29 1981-09-29 Character segmenting device of optical character reader

Publications (1)

Publication Number Publication Date
JPS5856079A true JPS5856079A (en) 1983-04-02

Family

ID=15576650

Family Applications (1)

Application Number Title Priority Date Filing Date
JP56154089A Pending JPS5856079A (en) 1981-09-29 1981-09-29 Character segmenting device of optical character reader

Country Status (1)

Country Link
JP (1) JPS5856079A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0424781A (en) * 1990-05-15 1992-01-28 Canon Inc Document processor
WO2013121647A1 (en) * 2012-02-17 2013-08-22 オムロン株式会社 Character-extraction method and character-recognition device and program using said method

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0424781A (en) * 1990-05-15 1992-01-28 Canon Inc Document processor
WO2013121647A1 (en) * 2012-02-17 2013-08-22 オムロン株式会社 Character-extraction method and character-recognition device and program using said method
JP2013171309A (en) * 2012-02-17 2013-09-02 Omron Corp Character segmentation method, and character recognition device and program using the same

Similar Documents

Publication Publication Date Title
US7684646B2 (en) System and method of determining image skew using connected components
US4375654A (en) Facsimile vector data compression
JP2766053B2 (en) Image data processing method
JP2003016440A (en) Photograph extraction method, device, program, and recording medium
US5608544A (en) Framed-area defining rectangle forming device
MXPA02008494A (en) Correction of distortions in form processing.
JPS5856079A (en) Character segmenting device of optical character reader
JPH1021316A (en) Mark reader and its method
JPH02130692A (en) Feature extracting circuit
JP2954218B2 (en) Image processing method and apparatus
JPH0214392A (en) Document area analyzing device
JP2980636B2 (en) Character recognition device
JP3275475B2 (en) Character string recognition device with known character sequence
JPS6343788B2 (en)
JPH05151350A (en) Method for correcting position distortion of image data
JP2679098B2 (en) Encoding processing device for contour detection image
JP2957050B2 (en) Image data enlarger
CN116521103A (en) Content printing method, device and medium for automatically acquiring printing information
JPH0820669B2 (en) Image information recording / reading method
JPH0442714B2 (en)
JPS59226978A (en) Skew correction system
JPH05250518A (en) Character recognizing method
JPS6292080A (en) Pattern recognizing device
JPH09261448A (en) Image processor
JPS59135573A (en) Identifying system of photo region