JPS62298887A - Character recognizing device - Google Patents

Character recognizing device

Info

Publication number
JPS62298887A
JPS62298887A JP61143389A JP14338986A JPS62298887A JP S62298887 A JPS62298887 A JP S62298887A JP 61143389 A JP61143389 A JP 61143389A JP 14338986 A JP14338986 A JP 14338986A JP S62298887 A JPS62298887 A JP S62298887A
Authority
JP
Japan
Prior art keywords
character string
character
storage means
range
recognition
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
JP61143389A
Other languages
Japanese (ja)
Other versions
JPH07120386B2 (en
Inventor
Hiroyuki Kami
上 博行
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.)
NEC Corp
Original Assignee
NEC Corp
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 NEC Corp filed Critical NEC Corp
Priority to JP61143389A priority Critical patent/JPH07120386B2/en
Publication of JPS62298887A publication Critical patent/JPS62298887A/en
Publication of JPH07120386B2 publication Critical patent/JPH07120386B2/en
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Landscapes

  • Character Input (AREA)

Abstract

PURPOSE:To eliminate the need for a large-capacity image storage means and a pressure compressing and expanding means by determining the objective range of recognition from a binary image obtained with low resolution and inputting a coordinate position within the range and resolution. CONSTITUTION:The binary image obtained with the low resolution is outputted by a photoelectric converting means 2 with a code for selecting the low resolution values of four points corresponding to the entire area, and the binary image is stored in an editing screen storage means 3. A display means 4 displays the binary image of the editing image storage means 3 and also displays a frame formed with the signal of the coordinates of the four points from an input means 5. The displayed frame is regarded as the objective range of recognition and when a code selecting resolution at the time of the scanning within the range is inputted from the input means 5, a binary image with high resolution is outputted from the photoelectric converting means 2 and stored in a character string image storage means 7. A character recognizing means 8 reads the image out of the character string image storage means 7 to perform character recognition.

Description

【発明の詳細な説明】 発明の詳細な説明 (産業上の利用分野) 本発明は編集用の画像を見ながら認識対象の範囲を指定
し、指定範囲内だけの文字認識を行う文字認識装置に関
する。
[Detailed Description of the Invention] Detailed Description of the Invention (Field of Industrial Application) The present invention relates to a character recognition device that specifies a recognition target range while viewing an editing image and performs character recognition only within the designated range. .

(i来の技術) 文字認識装置の光源変換部としては、−次元ラインセン
サに光を導く光学系を機械的に移動する方法が一般に採
用されている。編集用の画像と重ねてマウスやキーボー
ドなどの入力手段で制御される範囲を示す枠を表示しな
から認識対象領域を決め枠内の画像に対して文字認識を
行うときには、走査にかかる時間を短くするために、ま
づ1回の走査により求めた高い分解能の2値画像を高速
で読出しが可能な主記憶部あるいは容量の大きい補助記
憶部に記憶しておく。通常、表示部に表示出来るドツト
数に比較して記憶した画像のドツト数が多いので、つぎ
に記憶している画像をまびき得られる画像を編集画像と
し、表示された編集画像上での枠位置から原画像での位
置を求め、記憶されている原画像での文字列検出および
文字認識をおこなっていた。また記憶容量を減すために
画像を圧縮して記憶し、必要なとき伸張して使用する方
法もある。
(Previous technology) As a light source conversion unit of a character recognition device, a method is generally adopted in which an optical system that guides light to a -dimensional line sensor is mechanically moved. When displaying a frame indicating the range controlled by input means such as a mouse or keyboard over the image for editing and then determining the recognition target area and performing character recognition on the image within the frame, the time required for scanning is In order to shorten the time, first, a high-resolution binary image obtained by one scan is stored in a main storage section that can be read out at high speed or an auxiliary storage section that has a large capacity. Normally, the number of dots in the stored image is larger than the number of dots that can be displayed on the display, so the image obtained by combining the stored image is set as the edited image, and the frame position on the displayed edited image is The position in the original image was determined from the original image, and character string detection and character recognition were performed in the stored original image. There is also a method of compressing and storing images in order to reduce storage capacity, and decompressing and using them when necessary.

(発明が解決しようとする問題点) しかしながら前記方法では、最初から認識に不要な部分
も含めて走査・光電変換して得られた2値画像を記憶す
る必要があり、大容量の画像記憶手段を備えていなけれ
ばならない。圧縮伸張する方法では一般には特殊なハー
ドウェアでなされる圧縮伸張手段を必要とする。
(Problems to be Solved by the Invention) However, in the above method, it is necessary to store from the beginning a binary image obtained by scanning and photoelectric conversion, including parts unnecessary for recognition, and a large-capacity image storage means is required. must be prepared. Compression/expansion methods generally require compression/expansion means using special hardware.

本発明は、大きな容量の画像記憶手段と圧縮伸張手段と
が不要な文字認識装置の提供を目的とする。
An object of the present invention is to provide a character recognition device that does not require large-capacity image storage means and compression/expansion means.

(問題点を解決するための手段) 。(Means for solving problems).

本発明によれば、 (1)固定された帳票の指定範囲を指定の分解能で走査
・光電変換し2値画像を得る光電変換手段と、前記光電
変換手段からの低い分解能で得られた2値画像を記憶す
る編集画像記憶手段と、前記編集画像記憶手段の2値画
像と認識対象の範囲とを重ねて表示する表示手段と、前
記表示手段に表示される認識対象の範囲を決める座標位
置と分解能とを入力する入力手段と、前記入力手段で指
定された範囲に対応する前記編集画像記憶手段内の2値
画像より各文字列の上端、下端、左端及び右端の座標位
置を検出し各位置の座標値を出力する文字列位置検出手
段と、前記文字列位置検出手段からの各文字列ごとの範
囲を決める上下左右の4つの座標値を記憶する文字列位
置記憶手段と、前記文字列記憶手段の座標値を入力し前
記光電変換手段から出力される前記座標値で囲まれた範
囲内の2値画像を記憶する文字列画像記憶手段と、前記
文字列画像記憶手段からの2値画像に対して文字切り出
しと切り出された各画像に対しての文字認識とを行い認
識結果を出力する文字認識手段と、前記文字認識手段か
らの認識結果を記憶する認識結果記憶手段と、全体を制
御する制御手段とを備えだ文字認識装置と、 (2)固定された帳票の指定範囲を指定の分解能で走査
・光電変換し2値画像を得る光電変換手段と、前記光電
変換手段からの低い分解能で得られた2値画像を記憶す
る編集画像記憶手段と、前記編集画像記憶手段の2値画
像と認識対象の範囲とを重ねて表示する表示手段と、前
記表示手段に表示される認識対象の範囲を決める座標位
置と分解能とを入力する入力手段と、前記入力手段で指
定された範囲に対応する前記編集画像記憶手段内の2値
画像より各文字列の上端、下端、左端及び右端の座標位
置を検出し各位置の座標値を出力する文字列位置検出手
段と、前記文字列位置検出手段からの各文字列ごとの範
囲を決める上下左右の4つの座標値を記憶する文字列位
置記憶手段と、前記文字列位置記憶手段の各文字列の範
囲を決める座標値から作られる領域が前記文字列画像記
憶手段の記憶部に1回で入る文字列の組合せを求め最小
の組数における各組の領域を決める4つの座標値と各組
での各文字列の範囲を決める座標値とを出力する領域計
算手段と、前記文字列記憶手段の座標値を前記光電変換
手段に入力し出力される前記光電変換手段からこの前記
座標値で囲まれた範囲内の2値画像を記憶する文字列画
像記憶手段と、前記領域計算手段からの2値画像に対し
て前記文字列位置記憶手段からの各文字列の4つの座標
値により文字列の切出しと切出された文字列での文字切
り出し及び切り出された各画像に対しての文字認識とを
行い認識結果を出力する文字認識手段と、前記文字認識
手段からの認識結果を記憶する認識結果記憶手段と、全
体を制御する制御手段とを備えた文字認識装置とが得ら
れる。
According to the present invention, (1) a photoelectric conversion means that scans and photoelectrically converts a specified range of a fixed form at a specified resolution to obtain a binary image; and a binary image obtained from the photoelectric conversion means with a low resolution; an edited image storage means for storing an image; a display means for displaying a binary image of the edited image storage means and a recognition target range in an overlapping manner; and a coordinate position for determining the recognition target range displayed on the display means; an input means for inputting the resolution; and detecting the coordinate positions of the upper end, lower end, left end, and right end of each character string from the binary image in the edited image storage means corresponding to the range specified by the input means; a character string position detecting means for outputting the coordinate values of the character string position detecting means; a character string position storing means for storing four coordinate values of top, bottom, left and right that determine the range of each character string from the character string position detecting means; and the character string storing means. character string image storage means for inputting coordinate values of the means and storing a binary image within a range surrounded by the coordinate values output from the photoelectric conversion means; A character recognition means for cutting out characters and character recognition for each cut out image and outputting a recognition result, and a recognition result storage means for storing the recognition result from the character recognition means, and controlling the whole. (2) a character recognition device comprising a control means; (2) a photoelectric conversion means for scanning and photoelectrically converting a specified range of a fixed form at a specified resolution to obtain a binary image; an edited image storage means for storing the obtained binary image; a display means for displaying the binary image of the edited image storage means and a recognition target range in an overlapping manner; and a recognition target range displayed on the display means. an input means for inputting coordinate positions and resolution to determine the coordinate position, and coordinate positions of the upper end, lower end, left end, and right end of each character string from the binary image in the edited image storage means corresponding to the range specified by the input means; a character string position detecting means for detecting and outputting coordinate values of each position; and a character string position storing means for storing four coordinate values of top, bottom, left and right that determine the range of each character string from the character string position detecting means. , find the combinations of character strings in which the area created from the coordinate values that determine the range of each character string in the character string position storage means enters the storage section of the character string image storage means at one time; area calculation means for outputting four coordinate values that determine an area and coordinate values that determine the range of each character string in each set; character string image storage means for storing a binary image within a range surrounded by the coordinate values from the photoelectric conversion means; and character string image storage means for storing each character from the character string position storage means for the binary image from the area calculation means. a character recognition means for cutting out a character string based on four coordinate values of a column, character cutting out of the cut out character string, and character recognition for each cut out image, and outputting a recognition result; A character recognition device is obtained which includes a recognition result storage means for storing recognition results from the means and a control means for controlling the whole.

(作用) 光電変換部の走査位置が高い精度で制御でき固定されて
いる帳票自体を画像記憶手段とみなすと、特定領域にあ
る画像を読出し記憶することは光電変換部の走査位置を
制御し2値画像を得て記憶することに相当する。また文
字列の範囲を示す座標検出には、画像を水平と垂直との
方向に投影し得られるヒストグラムにいき値処理を施し
て求める方法が一般に使われているが、低い分解能の画
像からの投影情報からでも求めることができる。
(Function) If the scanning position of the photoelectric conversion unit can be controlled with high precision and the fixed form itself is regarded as an image storage means, reading and storing an image in a specific area requires controlling the scanning position of the photoelectric conversion unit. This corresponds to obtaining and storing a value image. In addition, to detect the coordinates that indicate the range of a character string, a method is generally used to calculate the histogram by projecting the image horizontally and vertically and applying threshold processing to the obtained histogram. It can also be obtained from information.

従って認識対象範囲を決める際には低い分解能の画像が
あればよく、帳票全部を高い分解能で入力し記憶してい
る必要はない。文字列の範囲を示す座標が得られると、
その座標値を用いて光電変換部の位置と分解能とを制御
し一文字列分の画像を入力する方法にすると、個々の文
字認識に用いる高い分解能での画像には少なくとも一文
字列分が記憶できる画像記憶手段を用意すればよい。
Therefore, when determining the recognition target range, it is sufficient to have an image with a low resolution, and it is not necessary to input and store the entire form at a high resolution. Once you have the coordinates that indicate the range of the string,
If the coordinate values are used to control the position and resolution of the photoelectric conversion unit and an image for one character string is input, the high-resolution image used for individual character recognition can store at least one character string. All you need to do is prepare a storage device.

−文字列置を記憶する画像記憶手段に必要な記憶容量は
帳票の最大幅と最大文字高さ及び光電変換部の分解能に
より決まるが、認識対象範囲の幅が帳票の幅と比較して
半分以下であれば複数文字列骨の画像を記憶することは
可能である。そこで予め検出された各文字列の範囲を組
み合わし画像記憶手段の記憶容量と分解能を条件に1回
の光電変換部の制御で取り込める範囲の組を求めておき
、範囲ごとに画像取り込みを行うと一文字分の画像を読
み出す時間を実質的に短縮できる。1回で取り込まれた
画像上での各文字列の範囲は範囲を決めるのに組み合わ
せた文字列の位置により計算できる。第3図は文字列か
ら作られる範囲の一例を説明11=へ するための図である。(a)の20は認識対象範囲、1
1から17までは認識対象範囲20内から検出された各
文字列の範囲を、また(b)は画像記憶手段に記憶でき
る容量に対応する文字列換算の範囲を表しているとする
。各文字列範囲の幅は範囲(b)の幅の半分以下である
ので、範囲(b)の面積以下の条件で文字列の範囲を組
み合わすと、範囲の組21,22.23が作れる。この
範囲の組ごとに光電変換部を制御すると、制御の回数は
3回ですみ、文字列ごとの制御回数7回より少ない。従
って前述のような機械的な制御が一般的である光電変換
部の制御回数が少なくなると、全体としての処理時間を
短縮できる。
- The storage capacity required for the image storage means that stores the character arrangement is determined by the maximum width and maximum character height of the form and the resolution of the photoelectric conversion unit, but the width of the recognition target range is less than half the width of the form. If so, it is possible to store images of multiple character string bones. Therefore, by combining the ranges of each character string detected in advance and finding a set of ranges that can be captured by controlling the photoelectric conversion unit once, subject to the storage capacity and resolution of the image storage means, image capture is performed for each range. The time required to read out an image for one character can be substantially reduced. The range of each character string on an image captured at one time can be calculated by the positions of the character strings that are combined to determine the range. FIG. 3 is a diagram for explaining an example of a range created from a character string to Explanation 11=. 20 in (a) is the recognition target range, 1
1 to 17 represent the range of each character string detected within the recognition target range 20, and (b) represents the range of character string conversion corresponding to the capacity that can be stored in the image storage means. Since the width of each character string range is less than half the width of range (b), by combining the character string ranges under the condition that the area is less than or equal to the area of range (b), range sets 21, 22, and 23 can be created. If the photoelectric conversion unit is controlled for each set in this range, the number of times of control is only three, which is less than the number of times of control for each character string, which is seven times. Therefore, by reducing the number of times the photoelectric conversion unit, which is generally mechanically controlled as described above, is controlled, the overall processing time can be shortened.

(実施例) 本発明を実施例を参照して詳細に説明する。(Example) The present invention will be explained in detail with reference to examples.

第1図は本発明の文字認識装置の一実施例を示すブロッ
ク図である。制御手段1から出力される分解能を選ぶ符
号と範囲を決める4点の座標値からなる制御信号101
により、光電変換手段2から2値画像の信号201と2
02とが出力される。制御手段1がら低い分解能を選ぶ
符号と全部の領域に相当する4点の座標値とを表す制御
信号101により光電変換手段2からは低い分解能で得
られた2値−画像が信号201として出力される。光電
変換手段2の走査部がCODに代表されるような線状の
光電変換素子、例えばCCDラインセンサ、に光を導く
光学系をステップモータにつなぎ移動させる形式では、
分解能の制御は光電変換素子からの映像信号に対するサ
ンプリング間隔とラインイメージを取込む間隔とを制御
することに相当する。編集画像記憶手段3は、制御手段
1からの制御信号102により光電変換手段2からの2
値画像を編集用の画像として記憶する。表示手段4は、
制御手段1からの制御信号103により編集画像記憶手
段3の2値画像を表示する。また表示手段4には入力手
段5からの4点の座標の信号501で作られる枠を表示
する。表示された枠内を認識対象の範囲とし、入力手段
5によりその範囲走査の際の分解能を選択する符号が入
力されると、その分解能選択の符号が信号502として
制御手段1に出力される。編集の際は見えるだけでよい
ので低い分解能の画像ですむが、文字認識の際には編集
で使った画像よりも高い分解能の画像が必要である。次
に制御手段1は座標の信号501を受けると、文字列位
置検出手段6に対して4点の座標値と文字列検出開始の
制御信号104を出力する。文字位置検出手段6は前記
制御信号104を入力すると、編集画像記憶手段3から
4点の座標自画像に対応する記憶している画像を読みだ
し文字列の検出を行う。文字列・の検出方法としては、
[スプリット検出法に基づく頁画像の構造解析」(辻、
浅井:電子通信学会技術研究報告、PRL85−17.
1985年6月21日)にあるように、画像の水平と垂
直方向の投影情報を利用する方法があり一般的であるの
でここでは文字列検出の詳細な説明は省略する。文字列
位置検出手段6からは、各文字列での範囲を決める4点
の座標値が順次、信号601として出力される。
FIG. 1 is a block diagram showing an embodiment of the character recognition device of the present invention. A control signal 101 consisting of a code for selecting the resolution and coordinate values of four points for determining the range, which is output from the control means 1.
As a result, binary image signals 201 and 2 are output from the photoelectric conversion means 2.
02 is output. A binary image obtained at a low resolution is output from the photoelectric conversion means 2 as a signal 201 in response to a control signal 101 representing a code for selecting a low resolution from the control means 1 and coordinate values of four points corresponding to the entire area. Ru. In a format in which the scanning section of the photoelectric conversion means 2 is connected to a step motor and moves an optical system that guides light to a linear photoelectric conversion element such as a COD, for example, a CCD line sensor,
Controlling the resolution corresponds to controlling the sampling interval for the video signal from the photoelectric conversion element and the interval at which line images are captured. The edited image storage means 3 receives the two images from the photoelectric conversion means 2 in accordance with the control signal 102 from the control means 1.
Store the value image as an image for editing. The display means 4 is
A control signal 103 from the control means 1 causes the binary image stored in the edited image storage means 3 to be displayed. Further, the display means 4 displays a frame formed by signals 501 of the coordinates of the four points from the input means 5. When the displayed frame is set as the range to be recognized and a code for selecting a resolution for scanning the range is inputted by the input means 5, the code for selecting the resolution is outputted to the control means 1 as a signal 502. When editing, you only need to see the image, so a low-resolution image is sufficient, but when character recognition requires an image with a higher resolution than the image used for editing. Next, when the control means 1 receives the coordinate signal 501, it outputs the coordinate values of the four points and a control signal 104 for starting character string detection to the character string position detection means 6. When the character position detection means 6 receives the control signal 104, it reads out the stored images corresponding to the four coordinate self-portraits from the edited image storage means 3 and detects a character string. As a method for detecting character strings,
[Structural analysis of page images based on split detection method] (Tsuji,
Asai: Institute of Electronics and Communication Engineers Technical Research Report, PRL85-17.
(June 21, 1985), there is a method that uses horizontal and vertical projection information of an image and is common, so a detailed explanation of character string detection will be omitted here. The character string position detection means 6 sequentially outputs the coordinate values of four points that determine the range of each character string as a signal 601.

制御手段1は入力される各文字列の4点の座標値である
信号601を記憶し、まず分解能を選ぶ信号502と最
初の1文字列の4点の座標値の信号601とを制御信号
101として出力する。光電変換手段2からは高い分解
能の2値画像の信号202が出力されるの靭 、乙(均 で、文字列画像記憶手段7は制御手段1からの記憶開始
の制御信号105により2値画像の信号202を記憶す
る。制御手段1は前記制御信号101により光電変換手
段2の1文字列分の走査を制御後に、文字認識開始の制
御信号106を文字認識手段8に出力する。
The control means 1 stores a signal 601 that is the coordinate values of four points of each input character string, and first outputs a signal 502 for selecting resolution and a signal 601 of the coordinate values of the four points of the first character string to the control signal 101. Output as . The photoelectric conversion means 2 outputs a binary image signal 202 with high resolution. A signal 202 is stored.The control means 1 controls the scanning of one character string by the photoelectric conversion means 2 using the control signal 101, and then outputs a control signal 106 for starting character recognition to the character recognition means 8.

文字認識手段8は前記制御信号106を入力すると、文
字列画像記憶手段7から1文字列分の画像を読みだし文
字切出しと切出された画像に対する文字認識を行う。文
字切出しの方法として、さまざまな方法が知られている
。たとえば[分散最小基準に基づく適応型文字分離方式
J(辻、浅井:電子通信学会論文誌、Vol、J68−
D、No、8.1985年8月)にあるような文字ピッ
チを推定し、推定ピッチをもとに最適化手法を利用して
最適な切出し位置を決定する方法があり、このような方
法を使うと文字切出しができる。また文字認識の方法も
多(の方法が知られている。ここでは適当な方法の1つ
たとえば特願昭60−270214 r文字認識方式j
の複数個の生別分析を使う方法を利用するとし、文字認
識方式の詳細な説明はここでは省略する。認識結果記憶
手段9は制御手段1からの前記制御信号106の後に出
力される記憶開始の制御信号107により文字認識手段
8からの認識結果を順次記憶する。認識結果記憶手段9
は文字認識手段8からの文字認識結果の出力が終了しそ
の結果の記憶が終わると終了信号901を制御手段1に
出力する。
When the character recognition means 8 receives the control signal 106, it reads out an image for one character string from the character string image storage means 7, cuts out characters, and performs character recognition on the cut out image. Various methods are known for character extraction. For example, [Adaptive character separation method J based on minimum variance criterion (Tsuji, Asai: Transactions of the Institute of Electronics and Communication Engineers, Vol. J68-
D, No. 8. August 1985), there is a method of estimating the character pitch and using an optimization method based on the estimated pitch to determine the optimal cutting position. You can use it to cut out characters. In addition, there are many known methods for character recognition.
A detailed explanation of the character recognition method will be omitted here, as a method using multiple bioanalyses will be used. The recognition result storage means 9 sequentially stores the recognition results from the character recognition means 8 in response to a storage start control signal 107 outputted after the control signal 106 from the control means 1. Recognition result storage means 9
outputs an end signal 901 to the control means 1 when the output of the character recognition result from the character recognition means 8 is completed and the storage of the result is completed.

制御手段1は信号901を入力すると、記憶している文
字列ごとに4点の座標値のうちで2番目の4点の座標値
と分解能を選ぶ符号とを制御信号101として光電変換
手段2に出力する。上記処理が繰り返されて、2番目の
文字列置の文字認識結果が認識結果記憶手段9に記憶さ
れる。同様にして、認識対象の範囲全体に対する文字認
識結果がもとまり、認識結果記憶手段9に記憶される。
When the control means 1 inputs the signal 901, it sends the coordinate values of the second four points among the coordinate values of the four points for each stored character string and the code for selecting the resolution to the photoelectric conversion means 2 as a control signal 101. Output. The above process is repeated and the character recognition result for the second character string position is stored in the recognition result storage means 9. Similarly, character recognition results for the entire recognition target range are collected and stored in the recognition result storage means 9.

文字列の画像を記憶できる容量が1文字列分の画像より
大きい場合には、前述のように、容量により1回の走査
部の制御で入れられる文字列の組みを求め各組の範囲を
決める4点の座標を記憶しておき、その座標値で光電変
換手段2を制御すると、頻繁に制御する必要がな(なり
画像の入力にかかる時間を短縮できる。
If the capacity to store an image of a character string is larger than the image for one character string, as described above, the range of each set is determined based on the capacity by determining the set of character strings that can be entered by controlling the scanning unit once. By storing the coordinates of the four points and controlling the photoelectric conversion means 2 using the coordinate values, frequent control is not necessary (and the time required to input an image can be shortened).

第2図は文字列画像を記憶できる容量が1文字列分より
大きい場合の本発明の文字認識装置の一実施例を示すブ
ロック図である。制御手段1から出力される分解能を選
ぶ符号と範囲を表す4点の座標値からなる制御信号10
1により、光電変換手段2から2値画像の信号201と
202とが出力される。制御手段1から低い分解能を選
ぶ符号と全部の領域に相当する4点の座標値とを表す制
御信号101により光電変換手段2からは低い分解能で
得られた2値画像が信号201として出力される。編集
画像記憶手段3は、制御手段1からの制御信号102に
より光電変換手段2からの2値画像を編集用の画像とし
て記憶する。
FIG. 2 is a block diagram showing an embodiment of the character recognition device of the present invention in which the storage capacity for character string images is larger than one character string. A control signal 10 consisting of a code for selecting resolution and coordinate values of four points representing a range outputted from the control means 1
1, the photoelectric conversion means 2 outputs binary image signals 201 and 202. A binary image obtained at a low resolution is output from the photoelectric conversion means 2 as a signal 201 in response to a control signal 101 representing a code for selecting a low resolution and coordinate values of four points corresponding to the entire area from the control means 1. . The edited image storage means 3 stores the binary image from the photoelectric conversion means 2 as an image for editing in accordance with the control signal 102 from the control means 1.

表示手段4は、制御手段1からの制御信号103により
編集画像記憶手段3の2値画像を表示する。また表示手
段4には入力手段5からの4点の座標の信号501で作
られる枠を表示する。表示された枠内が認識対象の範囲
として入力手段5によりその範囲走査の際の分解能を選
択する符号が入力され、信号jj’c3’ (柚 502として制御手段1に出力される。制御手段1は座
標の501を受けると、文字列位置検出手段6に対して
4点の座標値と文字列検出開始の信号104を出力する
。文字列位置検出手段6は信号104を入力すると、編
集画像記憶手段3から4点の座標内に対応する画像を読
みだし文字列の検出を行い、各文字列での範囲を決める
4点の座標値を順次、信号601として出力される。
The display means 4 displays the binary image stored in the edited image storage means 3 in response to the control signal 103 from the control means 1. Further, the display means 4 displays a frame formed by signals 501 of the coordinates of the four points from the input means 5. The displayed frame is the range to be recognized, and a code for selecting the resolution for scanning the range is input by the input means 5, and the signal jj'c3' (Yuzu 502) is output to the control means 1. When it receives the coordinate 501, it outputs the coordinate values of the four points and a signal 104 to start character string detection to the character string position detection means 6.When the character string position detection means 6 receives the signal 104, it outputs the edited image memory. Images corresponding to the coordinates of the four points are read out from the means 3, character strings are detected, and the coordinate values of the four points determining the range of each character string are sequentially output as a signal 601.

領域計算手段10は、まず制御手段1からの分解能を表
す信号105と文字列位置検出手段6からの各文字列の
4点の座標値である信号601とを記憶する。
The area calculation means 10 first stores a signal 105 representing the resolution from the control means 1 and a signal 601 representing the coordinate values of four points of each character string from the character string position detection means 6.

次に記憶した各文字列の範囲を決める4つの座標を組合
わせ、あらかじめ記憶している文字列画像記憶手段7の
容量と記憶した符号から決まる分解能とを条件に1回の
走査で取込める組合わせを求め、各組合せで作られた範
囲を表す4点の座標値と元の各文字列の範囲を決める4
点の座標値を信号1001として出力する。制御手段1
は前記信号1001を記憶し、まず分解能を選ぶ信号5
02と最初の組の4点の座標値の信号601とを制御信
号101として出力する。光電変換手段2からは高い分
解能の2値画像の信号202が出力されるので、文字列
画像記憶手段7は制御手段1からの制御信号106によ
り2値画像の信号202を記憶する。制御手段1は最初
の組の各文字列の4点の座標値の信号107を文字認識
手段8に出力する。文字認識子Fi8は前記信号107
を用いて文字列の切出し、文字切出し及び切出された画
像に対する文字認識を行う。認識結果記憶手段9は制御
手段1からの制御信号108により文字認識手段8から
の認識結果を順次記憶する。認識結果記憶手段9は文字
認識手段8からの文字認識結果の出力が終了しその結果
の記憶が終ると終了信号901を制御手段1に出力する
Next, the four coordinates that determine the range of each stored character string are combined to create a set that can be captured in one scan, subject to the capacity of the pre-stored character string image storage means 7 and the resolution determined from the stored code. Find the match and determine the coordinate values of the four points representing the range created by each combination and the range of each original character string 4
The coordinate value of the point is output as a signal 1001. Control means 1
stores the signal 1001 and first selects the resolution signal 5.
02 and a signal 601 of the coordinate values of the first set of four points are output as a control signal 101. Since the photoelectric conversion means 2 outputs a high-resolution binary image signal 202, the character string image storage means 7 stores the binary image signal 202 in accordance with the control signal 106 from the control means 1. The control means 1 outputs a signal 107 of the coordinate values of four points of each character string of the first set to the character recognition means 8. The character recognizer Fi8 uses the signal 107
is used to extract character strings, extract characters, and perform character recognition on the extracted images. The recognition result storage means 9 sequentially stores the recognition results from the character recognition means 8 in response to the control signal 108 from the control means 1. The recognition result storage means 9 outputs an end signal 901 to the control means 1 when the output of the character recognition result from the character recognition means 8 is completed and the storage of the result is completed.

制御手段1は前記信号901を入力すると、記憶してい
る組ごとの4点の座標値のうちで2番目の4点の座標値
と分解能を選ぶ符号とを制御信号101として光電変換
手段2に出力する。上記処理が繰り返されて、2番目の
組合の文字認識結果が認識結果記憶手段9に記憶される
。同様にして、認識対象の範囲全体に対する文字認識結
果がもとまり、認識結果記憶手段9に記憶される。
When the control means 1 receives the signal 901, it sends the coordinate values of the second four points among the stored coordinate values of the four points for each set and the code for selecting the resolution to the photoelectric conversion means 2 as a control signal 101. Output. The above process is repeated and the character recognition result of the second combination is stored in the recognition result storage means 9. Similarly, character recognition results for the entire recognition target range are collected and stored in the recognition result storage means 9.

上述の説明における手段は、メモリ、マイクロプロセッ
サ、ディスプレイ、キーボード(又はマウス)、スキャ
ナからなるパーソナルコンピュータシステムで行えるこ
とは言うまでもない。
It goes without saying that the means in the above description can be implemented in a personal computer system consisting of a memory, a microprocessor, a display, a keyboard (or mouse), and a scanner.

(発明の効果) 以上説明したように本発明によれば大きな容量の画像記
憶手段と圧縮伸張回路とが不要となる効果がある。
(Effects of the Invention) As explained above, according to the present invention, there is an effect that a large-capacity image storage means and a compression/expansion circuit are not required.

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

第1図、第2図は本発明の文字認識装置の一実施例を示
すブロック図、第3図は文字列から作られる範囲の一例
を示す図である。
1 and 2 are block diagrams showing an embodiment of the character recognition device of the present invention, and FIG. 3 is a diagram showing an example of a range created from a character string.

Claims (1)

【特許請求の範囲】 1、固定された帳票の指定範囲を指定の分解能で走査・
光電変換し2値画像を得る光電変換手段と、前記光電変
換手段からの低い分解能で得られた2値画像を記憶する
編集画像記憶手段と、前記編集画像記憶手段の2値画像
と認識対象の範囲とを重ねて表示する表示手段と、前記
表示手段に表示される認識対象の範囲を決める座標位置
と分解能とを入力する入力手段と、前記入力手段で指定
された範囲に対応する前記編集画像記憶手段内の2値画
像より各文字列の上端、下端、左端及び右端の座標位置
を検出し各位置の座標値を出力する文字列位置検出手段
と、前記文字列位置検出手段からの各文字列ごとの範囲
を決める上下左右の4つの座標値を記憶する文字列位置
記憶手段と、前記文字列記憶手段の座標値を入力し前記
光電変換手段から出力される前記座標値で囲まれた範囲
内の2値画像を記憶する文字列画像記憶手段と、前記文
字列画像記憶手段からの2値画像に対して文字切り出し
と切り出された各画像に対しての文字認識とを行い認識
結果を出力する文字認識手段と、前記文字認識手段から
の認識結果を記憶する認識結果記憶手段と、全体を制御
する制御手段とを備えることを特徴とする文字認識装置
。 2、固定された帳票の指定範囲を指定の分解能で走査・
光電変換し2値画像を得る光電変換手段と、前記光電変
換手段からの低い分解能で得られた2値画像を記憶する
編集画像記憶手段と、前記編集画像記憶手段の2値画像
と認識対象の範囲とを重ねて表示する表示手段と、前記
表示手段に表示される認識対象の範囲を決める座標位置
と分解能とを入力する入力手段と、前記入力手段で指定
された範囲に対応する前記編集画像記憶手段内の2値画
像より各文字列の上端、下端、左端及び右端の座標位置
を検出し各位置の座標値を出力する文字列位置検出手段
と、前記文字列位置検出手段からの各文字列ごとの範囲
を決める上下左右の4つの座標値を記憶する文字列位置
記憶手段と、前記文字列位置記憶手段の各文字列の範囲
を決める座標値から作られる領域が前記文字列画像記憶
手段の記憶部に1回で入る文字列の組合せを求め最小の
組数における各組の領域を決める4つの座標値と各組で
の各文字列の範囲を決める座標値とを出力する領域計算
手段と、前記文字列記憶手段の座標値を入力し前記領域
計算手段から出力される前記座標値で囲まれた範囲内の
2値画像を記憶する文字列画像記憶手段と、前記文字列
画像記憶手段がらの2値画像に対して前記文字列位置記
憶手段からの各文字列の4つの座標値により文字列の切
り出しと切り出された文字列での文字切り出し及び切り
出された各画像に対しての文字認識とを行い認識結果を
出力する文字認識手段と、前記文字認識手段からの認識
結果を記憶する認識結果記憶手段と、全体を制御する制
御手段とを備えることを特徴とする文字認識装置。
[Claims] 1. Scanning a specified range of a fixed form at a specified resolution.
A photoelectric conversion means that performs photoelectric conversion to obtain a binary image; an edited image storage means that stores the binary image obtained with low resolution from the photoelectric conversion means; a display means for displaying a range in an overlapping manner, an input means for inputting a coordinate position and resolution for determining a recognition target range displayed on the display means, and the edited image corresponding to the range specified by the input means. a character string position detecting means for detecting the coordinate positions of the upper end, lower end, left end, and right end of each character string from a binary image in the storage means and outputting the coordinate values of each position; and each character from the character string position detecting means. a character string position storage means for storing four coordinate values (up, down, left and right) that determine a range for each column; and a range surrounded by the coordinate values input from the character string storage means and output from the photoelectric conversion means. a character string image storage means for storing a binary image in the character string image storage means; character extraction is performed on the binary image from the character string image storage means, character recognition is performed on each of the extracted images, and a recognition result is outputted; What is claimed is: 1. A character recognition device comprising: a character recognition device that performs character recognition; a recognition result storage device that stores recognition results from the character recognition device; and a control device that controls the entire system. 2. Scan the specified range of a fixed form with the specified resolution.
A photoelectric conversion means that performs photoelectric conversion to obtain a binary image; an edited image storage means that stores the binary image obtained with low resolution from the photoelectric conversion means; a display means for displaying a range in an overlapping manner, an input means for inputting a coordinate position and resolution for determining a recognition target range displayed on the display means, and the edited image corresponding to the range specified by the input means. a character string position detecting means for detecting the coordinate positions of the upper end, lower end, left end, and right end of each character string from a binary image in the storage means and outputting the coordinate values of each position; and each character from the character string position detecting means. A character string position storage means for storing four coordinate values (up, down, left and right) that determine the range of each column, and an area created from the coordinate values that determine the range of each character string in the character string position storage means is the character string image storage means. Area calculation means for finding a combination of character strings that can be entered in the storage unit at one time and outputting four coordinate values that determine the area of each set in the minimum number of sets and coordinate values that determine the range of each character string in each set. a character string image storage means for inputting the coordinate values of the character string storage means and storing a binary image within a range surrounded by the coordinate values output from the area calculation means; and the character string image storage means Extracting character strings from the empty binary image using the four coordinate values of each character string from the character string position storage means, character extraction in the extracted character strings, and characters for each extracted image. 1. A character recognition device comprising: character recognition means for performing recognition and outputting a recognition result; recognition result storage means for storing recognition results from said character recognition means; and control means for controlling the whole.
JP61143389A 1986-06-18 1986-06-18 Character recognition device Expired - Lifetime JPH07120386B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP61143389A JPH07120386B2 (en) 1986-06-18 1986-06-18 Character recognition device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP61143389A JPH07120386B2 (en) 1986-06-18 1986-06-18 Character recognition device

Publications (2)

Publication Number Publication Date
JPS62298887A true JPS62298887A (en) 1987-12-25
JPH07120386B2 JPH07120386B2 (en) 1995-12-20

Family

ID=15337628

Family Applications (1)

Application Number Title Priority Date Filing Date
JP61143389A Expired - Lifetime JPH07120386B2 (en) 1986-06-18 1986-06-18 Character recognition device

Country Status (1)

Country Link
JP (1) JPH07120386B2 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH02187883A (en) * 1989-01-13 1990-07-24 Mitsubishi Electric Corp Document reader
US5361309A (en) * 1989-09-07 1994-11-01 Canon Kabushiki Kaisha Character recognition apparatus and method with low-resolution storage for character extraction

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH02187883A (en) * 1989-01-13 1990-07-24 Mitsubishi Electric Corp Document reader
US5361309A (en) * 1989-09-07 1994-11-01 Canon Kabushiki Kaisha Character recognition apparatus and method with low-resolution storage for character extraction

Also Published As

Publication number Publication date
JPH07120386B2 (en) 1995-12-20

Similar Documents

Publication Publication Date Title
US7949157B2 (en) Interpreting sign language gestures
US5187574A (en) Method for automatically adjusting field of view of television monitor system and apparatus for carrying out the same
US5048107A (en) Table region identification method
JP6095817B1 (en) Object detection device
US20100246968A1 (en) Image capturing apparatus, image processing method and recording medium
JP2849256B2 (en) Image recognition device
JPS62298887A (en) Character recognizing device
JP2002024762A (en) Document recognizing device and its method
US5563964A (en) Method and apparatus for processing a plurality of designated areas of an image
US5361309A (en) Character recognition apparatus and method with low-resolution storage for character extraction
JP2009301501A (en) Search result display method
JP3330348B2 (en) Video search method and apparatus, and recording medium storing video search program
WO2003063082A1 (en) Moving picture search apparatus
JP3421456B2 (en) Image processing device
JP2016129281A (en) Image processor
JP2803736B2 (en) Character recognition method
JPH0757044A (en) Character recognition device
JP3923104B2 (en) Table processing method and table processing apparatus
JP2926842B2 (en) Character extraction circuit
JPH01119885A (en) Document reader
JPH07168911A (en) Document recognition device
JPH0394393A (en) Character recognizing device
JPH0385681A (en) Picture processor
JPH08329198A (en) Document reader
JPS63189976A (en) Program chart input device