JPH03113687A - Character line detection system - Google Patents

Character line detection system

Info

Publication number
JPH03113687A
JPH03113687A JP1252842A JP25284289A JPH03113687A JP H03113687 A JPH03113687 A JP H03113687A JP 1252842 A JP1252842 A JP 1252842A JP 25284289 A JP25284289 A JP 25284289A JP H03113687 A JPH03113687 A JP H03113687A
Authority
JP
Japan
Prior art keywords
character
image data
data
vertical
character line
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
JP1252842A
Other languages
Japanese (ja)
Other versions
JP2725859B2 (en
Inventor
Akira Mori
晃 森
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.)
Sanyo Electric Co Ltd
Original Assignee
Sanyo Electric 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 Sanyo Electric Co Ltd filed Critical Sanyo Electric Co Ltd
Priority to JP1252842A priority Critical patent/JP2725859B2/en
Publication of JPH03113687A publication Critical patent/JPH03113687A/en
Application granted granted Critical
Publication of JP2725859B2 publication Critical patent/JP2725859B2/en
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Landscapes

  • Character Input (AREA)

Abstract

PURPOSE:To surely detect a character line containing a vertical ruled line or in a table frame by specifying a position where a black dot continues longer than a prescribed length by deciding the continuity of the black dot in a vertical direction in picture data, and detecting the character line as excluding data at this position or its neighbouring position. CONSTITUTION:A picture observing part 2, a picture memory 3, a character segmenting part 4, a standard pattern memory 6, and a recognizing part 5 are provided. Then, the position where the black dot continues longer than the prescribed length is specified by deciding the continuity of the black dot in the vertical direction in the picture data, and the horizontal projection of the picture data is generated as excluding the data of the specified position or its neighbouring position, and the character line of a recognized object is detected from the generated horizontal projection. Thus, character data in the table frame can be extracted by small memory capacity without generating the format of the table frame.

Description

【発明の詳細な説明】 (イ)産業上の利用分野 本発明は、文字認識装置において、認識対象の文字行を
検出する文字行検出方式に関する。
DETAILED DESCRIPTION OF THE INVENTION (A) Field of Industrial Application The present invention relates to a character line detection method for detecting a character line to be recognized in a character recognition device.

(ロ)従来の技術 一般に、文字認識装置においては、観測部により入力原
稿を走査光電変換して、黒ドツト及び白ドツトより成る
2値の画像データを得、該画像データを水平方向に走査
して水平射影を生成し、この水平射影から認識対象の文
字行を検出する。
(B) Conventional technology Generally, in a character recognition device, an input document is scanned and photoelectrically converted by an observation unit to obtain binary image data consisting of black dots and white dots, and this image data is scanned in the horizontal direction. A horizontal projection is generated, and a character line to be recognized is detected from this horizontal projection.

その後、検出諮れた文字行の画像データを垂直方向に走
査して垂直射影を生成し、この垂直射影に基づき文字の
切出しを行っていた。
Thereafter, the image data of the detected character line is scanned in the vertical direction to generate a vertical projection, and the character is extracted based on this vertical projection.

しかしながら、このような一般的な方式では、入力原稿
に文字と同様の読取り可能な色で表枠が記入されている
場合、表枠内の文字を切出すことはできず、従って、表
枠内の文字を認識することはできなかった。
However, with this general method, if the table frame is written in the input manuscript in the same readable color as the text, it is not possible to cut out the characters inside the table frame, and therefore, the characters inside the table frame cannot be cut out. I couldn't recognize the characters.

そこで、特開昭63−137383号公報に開示されて
いるように、認識に先立ち、表枠内に文字が記入されて
いない未記入原稿を先ず走査光電変換して、表枠のみの
画像データを予め記憶しておき、次に、表枠内に文字が
記入された入力原稿を走査光電変換し、得られた入力画
像データと記憶しておいた表枠画像データを比較するこ
とにより、表枠内の文字データを抽出する方式が、提案
啓れるようになった。
Therefore, as disclosed in Japanese Unexamined Patent Publication No. 137383/1983, prior to recognition, a blank document with no characters written in the front frame is first scanned and photoelectrically converted to convert image data of only the front frame. The table frame is stored in advance, and then the input document with characters written in the table frame is scanned and photoelectrically converted, and the obtained input image data is compared with the stored table frame image data. A method for extracting character data from within has been proposed.

(八)発明が解決しようとする課題 上記公報に開示された方式は、記入された表枠のフォー
マットが固定である場合には、有効であるが、新聞や文
献等のように、記入される表枠のフォーマットが固定で
ない場合には、わざわざ表枠のみの未記入原稿を作成し
なければならず、その作業が極めて煩雑であって、実用
性に乏しいものであった。
(8) Problems to be solved by the invention The method disclosed in the above gazette is effective when the format of the table frame filled in is fixed, but as in newspapers, literature, etc. If the format of the table frame is not fixed, a blank manuscript with only the table frame must be created, which is extremely complicated and impractical.

更に、上記方式では、入力原稿の画像データと未記入原
稿の画像データとを比較することから、画像データを記
憶するためのメモリ容量が増加し、且つ、処理が複雑に
なるという課題があった。
Furthermore, in the above method, since the image data of the input document is compared with the image data of the blank document, there is a problem that the memory capacity for storing the image data increases and the processing becomes complicated. .

〈二)課題を解決するための手段 本発明は、文字と共に読取り可能な縦罫線が記入された
原稿を走査光電変換し、得られた画像データに基づき文
字の認識を行う装置において、前記画像データにおける
黒ドツトの垂直方向の連続性を判定して、所定長以上黒
ドツトが連続する位置を特定し、該特定された位置もし
くはその近傍位置のデータを除いて前記画像データの水
平射影を生成し、該生成された水平射影から認識対象の
文字行を検出することにより、上記課題を解決するもの
である。
(2) Means for Solving the Problems The present invention provides an apparatus for scanning and photoelectrically converting a document on which vertical ruled lines that can be read along with characters are written, and recognizing characters based on the obtained image data. determine the vertical continuity of black dots in the image data, identify positions where black dots are continuous for a predetermined length or longer, and generate a horizontal projection of the image data by excluding data at the identified positions or positions in the vicinity thereof. , the above problem is solved by detecting the character line to be recognized from the generated horizontal projection.

又、本発明は、縦罫線及び横罫線より成る読取り可能な
表枠内に文字が記入された原稿を、走査光電変換して画
像データを得、得られた画像データに基づき文字の認識
を行う装置において、前記画像データにおける黒ドツト
の垂直方向の連続性を判定して、所定長以上黒ドツトが
連続する位置を特定し、該特定された位置もしくはその
近傍位置のデータを除いて前記画像データの水平射影を
生成し、該生成された水平射影のうち、有効長が所定長
以上の水平射影から認識対象の文字行を検出することに
より、上記課題を解決するものである。
Further, the present invention performs scanning photoelectric conversion of a document in which characters are written in a readable table frame consisting of vertical ruled lines and horizontal ruled lines to obtain image data, and recognizes the characters based on the obtained image data. The apparatus determines the vertical continuity of black dots in the image data, identifies positions where black dots are continuous for a predetermined length or more, and removes data at the identified positions or positions in the vicinity thereof and processes the image data. The above problem is solved by generating a horizontal projection of , and detecting a character line to be recognized from among the generated horizontal projections, the effective length of which is greater than or equal to a predetermined length.

尚、ここで、黒ドツトとは、記入された罫線及び文字に
対応する画像データを示し、白ドツトは空白に対応する
Note that here, the black dots indicate image data corresponding to written ruled lines and characters, and the white dots correspond to blank spaces.

(*)作用 本発明では、画像データにおいて所定長以上黒ドツトが
連続する位置が特定きれることにより、画像データ中の
縦罫線の位置が検出され、この位置もしくはその近傍位
置のデータを除いて画像データの水平射影が生成きれる
ため、縦罫線データもしくはその近傍データが除外され
た文字のみのデータについて水平射影が生成され、従っ
て、この水平射影から文字行の検出が可能となる。
(*) Effect In the present invention, the position of a vertical ruled line in the image data is detected by identifying the position in the image data where black dots are continuous for a predetermined length or more, and the data at this position or its vicinity is excluded from the image. Since a horizontal projection of data can be generated, a horizontal projection is generated for only character data from which vertical ruled line data or data in the vicinity thereof has been excluded, and character lines can therefore be detected from this horizontal projection.

又、本発明では、横罫線の水平射影の有効長が所定長以
下になることを考慮して、更に、上記生成された水平射
影のうちその有効長が所定長以上の水平射影から文字行
を検出するので、横罫線を含む場合もその横罫線は除外
され、従って、表枠内の文字のみのデータから成る文字
行が検出可能となる。
Furthermore, in the present invention, considering that the effective length of the horizontal projection of the horizontal ruled line is less than or equal to a predetermined length, character lines are further extracted from the horizontal projections whose effective length is greater than or equal to the predetermined length among the horizontal projections generated above. Since the table frame is detected, even if a horizontal ruled line is included, the horizontal ruled line is excluded, and therefore a character line consisting of data of only characters within the table frame can be detected.

(へ)実施例 第2図は、本発明の方式を実現する文字認識装置の構成
を示すブロック図であり、(1,)は縦罫線及び横罫線
より成る読取り可能な表枠内に文字が印刷された原稿、
(2)は原稿(1)を走査光電変換して画像データを得
る画像観測部、(3)は画像観測部(2)により得られ
た画像データを記憶する画像メモリ、(4)は画像メモ
リ(3)から認識対象文字を一文字づつ切出す文字切出
し部、(5)は切出された文字パターンから特徴パター
ンを抽出し、抽出した特徴パターンを、標準パターンメ
モリ〈6)に予め記憶された標準パターンと照合して、
認識を行う認識部である、そして、この文字切出し部(
4)は、文字行検出部(7)、縦罫線検出部(8〉、文
字行再検出部(9)、文字切出し制御部(10)より構
成されている。
(v) Embodiment FIG. 2 is a block diagram showing the configuration of a character recognition device that implements the method of the present invention. printed manuscript,
(2) is an image observation section that scans and photoelectrically converts the original (1) to obtain image data; (3) is an image memory that stores the image data obtained by the image observation section (2); and (4) is an image memory. (3) A character extraction unit that extracts the recognition target characters one by one from the character pattern, (5) extracts a feature pattern from the extracted character pattern, and stores the extracted feature pattern in the standard pattern memory (6) in advance. By comparing it with the standard pattern,
This is the recognition unit that performs recognition, and this character extraction unit (
4) is composed of a character line detection section (7), a vertical ruled line detection section (8>), a character line redetection section (9), and a character cutout control section (10).

以下、第1図のフローチャート及び第3図〜第5図を参
照して、文字切出し部(4)の処理内容を詳しく説明す
る。
Hereinafter, the processing contents of the character cutting section (4) will be explained in detail with reference to the flowchart of FIG. 1 and FIGS. 3 to 5.

ここでは、原稿(1)には、第3図に示すように、表枠
(11)内に文字が印刷され、且つ、表枠外の第1行目
にも文字が印刷されているものとする。
Here, it is assumed that in the manuscript (1), as shown in Figure 3, characters are printed inside the front frame (11), and characters are also printed on the first line outside the front frame. .

さて、原稿(1)の画像データが画像メモリ(3)に記
憶されると、文字切出し部<4)内の文字行検出部(7
)が先ず動作し、画像データを水平方向に走査して、各
水平ドツトライン毎に、ライン内の全ドツトをOR演算
することにより、水平射影(12)を生成する。この水
平射影(12)の生成は、水平射影の点部分の長さ(以
下、有効長という)Lが、所定長り。に達するか、もし
くは、達する以前に途切れるまで実行され、本実施例で
は、この所定長L0を、文字サイズNの2゜5倍の2.
5Nに設定している。そして、文字行検出部(7)は、
この水平射影の有効長しが所定長り。に達したときは、
文字行検出不能と判断し、所定長り。より短かければ文
字行検出可能と判断する。
Now, when the image data of the original (1) is stored in the image memory (3), the character line detection part (7) in the character cutting part <4) is stored in the image memory (3).
) is first operated to horizontally scan the image data and produce, for each horizontal dot line, a horizontal projection (12) by ORing all dots in the line. In the generation of this horizontal projection (12), the length (hereinafter referred to as effective length) L of the point portion of the horizontal projection is a predetermined length. In this embodiment, this predetermined length L0 is set to 2.5 times the character size N, which is 2.5 times the character size N.
It is set to 5N. Then, the character line detection unit (7)
The effective length of this horizontal projection is the predetermined length. When it reaches
It is determined that the character line cannot be detected, and the specified length is determined. If it is shorter, it is determined that a character line can be detected.

第3図の原稿(1)の第1行目には、文字のみが印刷さ
れているため、その水平射影(12a)の有効長りはN
以下となって、所定長L0より短かくなり、文字行検出
可能と判断される。このような場合は、従来と同様、水
平射影(12g)の先端(100)と後端(101)の
垂直座標で囲まれた画像データを、認識対象の文字行と
して検出し、文字切出し制御部(10)に処理が移る。
Since only characters are printed on the first line of the manuscript (1) in Figure 3, the effective length of its horizontal projection (12a) is N
The length is shorter than the predetermined length L0, and it is determined that the character line can be detected. In such a case, as in the past, the image data surrounded by the vertical coordinates of the leading edge (100) and trailing edge (101) of the horizontal projection (12g) is detected as a character line to be recognized, and the character cutting control unit The process moves to (10).

この制御部(10)では、第4図に示すように、検出し
た文字行(13)中の画像データを垂直方向に走査し、
各垂直ドツトライン毎に、ライン内の全ドツトをOR演
算することにより、垂直射影(14)を生成し、この垂
直射影(14)に基づき文字の切出しを行う。
As shown in FIG. 4, this control unit (10) scans the image data in the detected character line (13) in the vertical direction,
For each vertical dot line, a vertical projection (14) is generated by ORing all the dots in the line, and characters are extracted based on this vertical projection (14).

一方、1行目の文字の切出し後、次の画像データについ
て水平射影を生成すると、縦罫線の存在により、水平射
影(12b )の有効長りは第3図に示すように、所定
長L0に達してしまい、文字行検出不能と判断される。
On the other hand, when a horizontal projection is generated for the next image data after cutting out the characters in the first line, the effective length of the horizontal projection (12b) is reduced to a predetermined length L0 due to the presence of vertical ruled lines, as shown in Figure 3. It is determined that the character line cannot be detected.

この場合、文字行検出部(7)は処理を縦罫線検出部(
8)へ移す。
In this case, the character line detection unit (7) performs the processing by the vertical ruled line detection unit (7).
Move to 8).

縦罫線検出部(8)では、文字行検出不能と判断された
領域(15)内の画像データに対し、黒ドツトの垂直方
向の連続性を判定して、黒ドツトが所定長L0連続する
データの水平位置を特定する。具体的には、領域(15
)内の画像データを垂直方向に走査して、各垂直ドツト
ライン毎に、ライン内の全ドツトをAND演算し、この
AND結果がrl」である垂直ドツトラインの水平座標
を記憶する。以下、このAND演算により得られた結果
を垂直積算射影と呼び、第5図においては、(16)で
示している。つまり、この垂直積算射影(16)は、文
字成分を取り除いて表枠(11)を構成する縦罫線の成
分のみを抽出した結果であり、従って、この垂直積算射
影(16)の存在する位置により、縦罫線の位置が検出
される。
The vertical ruled line detection unit (8) determines the vertical continuity of black dots in the image data in the area (15) in which it is determined that character lines cannot be detected, and detects data in which black dots continue for a predetermined length L0. Determine the horizontal position of Specifically, the area (15
) is scanned in the vertical direction, and for each vertical dot line, all the dots in the line are ANDed, and the horizontal coordinate of the vertical dot line whose AND result is rl is stored. Hereinafter, the result obtained by this AND operation will be referred to as a vertical integration projection, and is indicated by (16) in FIG. In other words, this vertical cumulative projection (16) is the result of removing the character components and extracting only the vertical ruled line components that make up the table frame (11). , the position of the vertical ruled line is detected.

縦罫線位置の検出後、処理は文字行再切出部(9〉に移
り、領域(15)の画像データに対し、水平射影の再生
成が行われる。但し、この場合は、縦罫線検出部(8)
で検出された位置情報に基づき、第5図斜線で示す縦罫
線及びその近傍領域(17a)(17b)(17c)(
17d)内のデータを除いて、水平射影が生成きれるも
のであり、従って、この処理により、表枠(11)内の
文字及び横罫線の射影が、第5図に示すように浮かび上
がる。更に、文字行再切出部(9)は、再生成された水
平射影の有効長が所定長り、(例えば、文字サイズNの
175程度)より長いか否か判定し、所定長し□以上の
水平射影のみを文字行の水平射影とする。これにより文
字に比べて有効長がかなり短かい横罫線の水平射影は無
視されることとなる。
After detecting the vertical ruled line position, the process moves to the character line re-extraction section (9>), and the horizontal projection is regenerated for the image data of the area (15).However, in this case, the vertical ruled line detection section (8)
Based on the position information detected in , vertical ruled lines and their neighboring areas (17a) (17b) (17c) (
Horizontal projections can be generated except for the data in 17d), and therefore, by this process, the projections of the characters and horizontal ruled lines in the table frame (11) emerge as shown in FIG. Furthermore, the character line re-cutting unit (9) determines whether the effective length of the regenerated horizontal projection is longer than a predetermined length (for example, about 175 of the character size N), and determines whether or not the effective length of the regenerated horizontal projection is longer than □ Let only the horizontal projection of be the horizontal projection of a character line. As a result, the horizontal projection of horizontal ruled lines, whose effective length is much shorter than that of characters, is ignored.

依って、第5図の場合、横罫線(18)の水平射影(1
2c)は無視され、次の水平射影(12d)の先端(1
02)と後端(103)の垂直座標により囲まれた画像
データが、表枠(11)内の第1行の文字行として検出
される。その後、処理は文字切出し制御部(10)に移
り、上述した垂直射影の生成により、文字の切出しが実
行される。
Therefore, in the case of Fig. 5, the horizontal projection (1) of the horizontal ruled line (18) is
2c) is ignored and the tip (1
The image data surrounded by the vertical coordinates of 02) and the rear end (103) is detected as the first character line in the table frame (11). Thereafter, the process moves to the character segmentation control unit (10), where character segmentation is executed by generating the above-described vertical projection.

以下、検出された文字行(19)以降の画像データにつ
いて、第1図フローチャートに示す同様の処理を繰り返
せば、表枠(11)内の第2行、第3行。
Hereinafter, if the same process shown in the flowchart of FIG. 1 is repeated for the image data after the detected character line (19), the second and third lines in the table frame (11) will be obtained.

・・・・・・・・・最終行についても、順次文字行の検
出及び文字行内の文字切出しが可能となる。
. . . Even for the last line, it is possible to sequentially detect character lines and cut out characters within the character line.

(ト)発明の効果 本発明に依れば、認識原稿を観測部に入力するだけで、
縦罫線を含む文字行あるいは表枠内の文字行を確実に検
出でき、これら文字行内の文字の認識が可能となる。従
って、表枠のみの未記入原稿の作成及び入力という煩し
い作業は一切不要となり、フォーマットが固定されてい
ない印刷物の文字認識においては、極めて効果が大なる
ものである。
(g) Effects of the invention According to the present invention, by simply inputting a recognized manuscript into the observation section,
Character lines including vertical ruled lines or character lines within a table frame can be reliably detected, and characters within these character lines can be recognized. Therefore, the troublesome work of creating and inputting a blank manuscript with only a front frame is no longer necessary, and the present invention is extremely effective in character recognition of printed matter whose format is not fixed.

更に、未記入原稿画像データと認識原稿画像データの比
較を行わないため、処理が簡単になると共に、メモリ容
量を増加させる必要がなくなる。
Furthermore, since no comparison is made between blank document image data and recognized document image data, processing is simplified and there is no need to increase memory capacity.

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

第1図は本発明の実施例の処理内容を示すフローチャー
ト、第2図は実施例の構成を示すブロック図、第3図乃
至第5図は実施例の処理内容を説明するための説明図で
ある。 (1)・・・原稿、 (2)・・・画像観測部、 (3
)・・・画像メモリ、 〈4〉・・・文字切出し部、 
(7)・・・文字行検出部、 (8)・・・縦罫線検出
部、 (9)・・・文字行再検出部、 (10)・・・
文字切出し制御部。
FIG. 1 is a flowchart showing the processing content of an embodiment of the present invention, FIG. 2 is a block diagram showing the configuration of the embodiment, and FIGS. 3 to 5 are explanatory diagrams for explaining the processing content of the embodiment. be. (1)... Manuscript, (2)... Image observation department, (3
)...Image memory, <4>...Character cutting section,
(7)...Character line detection unit, (8)...Vertical ruled line detection unit, (9)...Character line redetection unit, (10)...
Character extraction control section.

Claims (2)

【特許請求の範囲】[Claims] (1)文字と共に読取り可能な縦罫線が記入された原稿
を走査光電変換し、得られた画像データに基づき文字の
認識を行う装置において、前記画像データにおける黒ド
ットの垂直方向の連続性を判定して、所定長以上黒ドッ
トが連続する位置を特定し、該特定された位置もしくは
その近傍位置のデータを除いて前記画像データの水平射
影を生成し、該生成された水平射影から認識対象の文字
行を検出することを特徴とした文字行検出方式。
(1) In a device that scans and photoelectrically converts a document with vertical ruled lines that can be read along with characters, and recognizes characters based on the obtained image data, determines the vertical continuity of black dots in the image data. Then, identify a position where black dots are continuous for a predetermined length or longer, generate a horizontal projection of the image data by excluding data at the identified position or its vicinity, and calculate the recognition target from the generated horizontal projection. A character line detection method characterized by detecting character lines.
(2)縦罫線及び横罫線より成る読取り可能な表枠内に
文字が記入された原稿を、走査光電変換して画像データ
を得、得られた画像データに基づき文字の認識を行う装
置において、前記画像データにおける黒ドットの垂直方
向の連続性を判定して、所定長以上黒ドットが連続する
位置を特定し、該特定された位置もしくはその近傍位置
のデータを除いて前記画像データの水平射影を生成し、
該生成された水平射影のうち有効長が所定長以上の水平
射影から認識対象の文字行を検出することを特徴とした
文字行検出方式。
(2) In a device that scans and photoelectrically converts a document in which characters are written in a readable table frame consisting of vertical and horizontal ruled lines to obtain image data, and recognizes the characters based on the obtained image data, The vertical continuity of black dots in the image data is determined, a position where black dots are continuous for a predetermined length or more is specified, and the horizontal projection of the image data is performed by excluding data at the specified position or a position in its vicinity. generate,
A character line detection method characterized in that a character line to be recognized is detected from a horizontal projection whose effective length is a predetermined length or more among the generated horizontal projections.
JP1252842A 1989-09-28 1989-09-28 Character line detection method Expired - Lifetime JP2725859B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP1252842A JP2725859B2 (en) 1989-09-28 1989-09-28 Character line detection method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP1252842A JP2725859B2 (en) 1989-09-28 1989-09-28 Character line detection method

Publications (2)

Publication Number Publication Date
JPH03113687A true JPH03113687A (en) 1991-05-15
JP2725859B2 JP2725859B2 (en) 1998-03-11

Family

ID=17242943

Family Applications (1)

Application Number Title Priority Date Filing Date
JP1252842A Expired - Lifetime JP2725859B2 (en) 1989-09-28 1989-09-28 Character line detection method

Country Status (1)

Country Link
JP (1) JP2725859B2 (en)

Also Published As

Publication number Publication date
JP2725859B2 (en) 1998-03-11

Similar Documents

Publication Publication Date Title
US5075895A (en) Method and apparatus for recognizing table area formed in binary image of document
JP2940936B2 (en) Tablespace identification method
JP4189506B2 (en) Apparatus, method and recording medium for image processing
JP4655335B2 (en) Image recognition apparatus, image recognition method, and computer-readable recording medium on which image recognition program is recorded
US8229214B2 (en) Image processing apparatus and image processing method
JP2002015280A (en) Device and method for image recognition, and computer- readable recording medium with recorded image recognizing program
JP3858559B2 (en) Image processing apparatus, image processing method, and recording medium
JPH0991453A (en) Image processing method and its unit
JPH03113687A (en) Character line detection system
JP2800192B2 (en) High-speed character / graphic separation device
JP2851087B2 (en) Table processing method
JP2923004B2 (en) Image generation device
JP2579356B2 (en) Document reading device
JP2002342880A (en) Vehicle number recognition device
JPS6327752B2 (en)
JPS63157287A (en) Character pattern extracting method
JPH0728934A (en) Document image processor
JP2023034823A (en) Image processing apparatus, and control method, and program for image processing apparatus
JP4230478B2 (en) Document processing apparatus, method and program
JP3081510B2 (en) Braille recognition system
JPH05174179A (en) Document image processor
JP2003123076A (en) Image processor and image processing program
JPH03160582A (en) Method for separating ruled line and character in document picture data
JPH0271379A (en) Picture processor
JPH02166583A (en) Character recognizing device