JPS5941232B2 - character recognition device - Google Patents
character recognition deviceInfo
- Publication number
- JPS5941232B2 JPS5941232B2 JP54135613A JP13561379A JPS5941232B2 JP S5941232 B2 JPS5941232 B2 JP S5941232B2 JP 54135613 A JP54135613 A JP 54135613A JP 13561379 A JP13561379 A JP 13561379A JP S5941232 B2 JPS5941232 B2 JP S5941232B2
- Authority
- JP
- Japan
- Prior art keywords
- character
- judgment
- circuit
- register
- determination
- 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
Links
Description
【発明の詳細な説明】
本発明は文字認識が簡単な構成で容易に行なえるように
した文字認識装置に関するものである。DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a character recognition device that allows character recognition to be easily performed with a simple configuration.
従来のこの種の装置においては2次判定で新たに必要な
特徴を、1次判定処理で得たコード化列や文字パターン
から抽出する処理が必要であり、多大な処理時間がかか
つたり、装置のコストが高くなる等の欠点があつた。本
発明はこのような欠点を除去するもので、以下にその実
施例とともに説明する。In conventional devices of this type, it is necessary to extract newly necessary features in the secondary judgment from the coded strings and character patterns obtained in the primary judgment process, which takes a large amount of processing time. There were drawbacks such as increased cost of the device. The present invention eliminates these drawbacks and will be explained below along with examples thereof.
第1図において1は帳票等に記載された読取対象文字を
走査し、光電変換し、2値化信号を得る回路、2は回路
1で得られた文字パターンの2値化信号を記憶するメモ
リ、3はメモリ2の内容にもとづき、基準点(たとえば
文字の最下端)から一定廻り(たとえば時計廻り)に文
字輪郭をトレースし、輪郭の単位長さごとに方向コード
化し(たとえば0〜7のコード)、メモリ2に記憶情報
をその文字の輪郭の方向コード列情報に変換する変換回
路である。4は変換回路で得られた文字輪郭方向コード
列を記憶するメモリ、5は読取対象全文字カテゴリーの
標準パターンが記憶されているメモリである。In FIG. 1, 1 is a circuit that scans the characters to be read written on a form, photoelectrically converts them, and obtains a binary signal, and 2 is a memory that stores the binary signal of the character pattern obtained in circuit 1. , 3 traces the character outline in a fixed direction (e.g. clockwise) from the reference point (e.g. the bottom edge of the character) based on the contents of the memory 2, and codes the direction for each unit length of the outline (e.g. 0 to 7). This is a conversion circuit that converts the information stored in the memory 2 into the direction code string information of the outline of the character. Reference numeral 4 indicates a memory for storing the character outline direction code string obtained by the conversion circuit, and reference numeral 5 indicates a memory for storing standard patterns of all character categories to be read.
6は1次判定回路で、全文字分の標準パターンの中から
1文字カテゴリーずつ全文字カテゴリーと参照し、各文
字カテゴリーごとにその標準パターンで指定された線分
区切り情報(例えば上記方向コード列で、一定方向が一
定数続いた時線分区切りとする)にもとづき上記方向コ
ード列を線分に区切り、1つ1つの線分について標準パ
ターンに記憶されている線分特徴(例えば長さ、凹凸具
合等)と上記方向コード列の適合性を比較する。Reference numeral 6 denotes a primary judgment circuit, which refers to all character categories one by one from among the standard patterns for all characters, and uses the line segment separation information (for example, the above direction code string) specified by the standard pattern for each character category. The above direction code string is divided into line segments based on a certain number of consecutive directions, and the line segment characteristics (for example, length, Compare the degree of unevenness, etc.) and the suitability of the above direction code string.
この場合、1つの文字カテゴリーの標準バターンについ
て上記線分に分割された方向コード列がすべて適合して
いればその文字カテゴリーを判定候補文字カテゴリーと
判断し、判定候補カテゴリーレジスタ7にセツトする。
さらに1次判定回路6は上記標準パターンにもとづく線
分の分割及び特徴の比較時に、該当標準パターン線分情
報と共に標準パターンメモリ5内に記憶されている文字
特徴レジスタセツト指令にもとづきその線分と次の線分
の区切り点のX,Y軸座標位置及びその点までの方向コ
ード数その他指令された情報を指令された文字特徴レジ
スタ8にセツトする。9は上記レジスタ7に記憶された
候補文字カテゴリーごとに、上記レジスタ8に記憶され
た文字特徴情報にもとづき2次判定(最終文字判定)を
行なう判定回路である。In this case, if all the direction code strings divided into line segments match the standard pattern of one character category, that character category is determined to be a judgment candidate character category, and is set in the judgment candidate category register 7.
Furthermore, when dividing the line segment based on the standard pattern and comparing the features, the primary judgment circuit 6 divides the line segment based on the character feature register set command stored in the standard pattern memory 5 together with the corresponding standard pattern line segment information. The X and Y axis coordinate position of the break point of the next line segment, the number of direction codes up to that point, and other commanded information are set in the commanded character feature register 8. A determination circuit 9 performs a secondary determination (final character determination) for each candidate character category stored in the register 7, based on the character feature information stored in the register 8.
次に上記実施例の動作について説明する。Next, the operation of the above embodiment will be explained.
たとえばメモリ5に記載された文字カテゴリ゛J゛の標
準パターンには第2図イのA−G点で線分に区切られる
べく線分区切り情報が格納されている。またA点を基準
としたB点およびE点のY軸座標が文字特徴レジスタ8
の1番地および2番地にセツトされるべくセツト指令が
格納されている。今、文字特徴レジスタ8の1番地にセ
ツトされた内容をy1、文字特徴レジスタ8の番地にセ
ツトさY2−y1れた内容をY2とした時、2次判定で
は
Y2
を計算し、文字カテゴリーJとしてこの値が適当か否か
を判定する。For example, in the standard pattern of the character category "J" written in the memory 5, line segment separation information is stored so that the pattern is separated into line segments at points A-G in FIG. 2A. In addition, the Y-axis coordinates of point B and point E with respect to point A are in the character feature register 8.
Set commands to be set are stored at addresses 1 and 2 of . Now, if the content set at address 1 of the character feature register 8 is y1, and the content set at address Y2-y1 of the character feature register 8 is set as Y2, then in the secondary judgment Y2 is calculated and the character category J to determine whether this value is appropriate.
この判定は標準パターンとの線分特徴の照合(これを1
次判定という)のみでは第2図の帽こ示すような゛゛と
の区別が明確でないために行なわれるものである。この
様に2次判定は、1次判定で行なわれる画一的処理に対
し、1次判定で候補として挙げられた文字1つ1つにつ
いて、前記文字特徴レジスタ8の内容を使い個別にその
文字カテゴリーの特徴を柔軟に判定するものであり、2
次判定処理自身はマイクロコンピユータ等を使つたスト
アトプログラム方式をとるのが適当である。以上の実施
例より明らかなように本発明によれば1次判定処理過程
内でレジスタセツトされたX,Y座標または方向コード
列の位置情報を2次判定処理でそのまま用いるため、2
次判定で新たに必要な特徴を前記コード化列や、文字発
生パターンから抽出する処理が不必要になり、処理時間
及びハードウエアコストの低減につながる利点がある。This judgment is performed by matching the line segment features with the standard pattern (this is
This is done because it is not clear to distinguish between `` and '' as shown in the cap in FIG. In this way, in contrast to the uniform processing performed in the primary determination, the secondary determination uses the contents of the character feature register 8 to individually evaluate each character selected as a candidate in the primary determination. It is a method that flexibly determines the characteristics of a category, and 2.
It is appropriate that the next determination process itself takes the form of a stored program using a microcomputer or the like. As is clear from the above embodiments, according to the present invention, the position information of the X, Y coordinates or direction code string set in the register during the primary determination process is used as is in the secondary determination process.
This eliminates the need for the process of extracting newly required features from the coded string or character generation pattern in the next determination, which has the advantage of reducing processing time and hardware costs.
また、本発明は、読み取つた図形を1次判定回路で判定
して、文字を特定できなかつた場合でも、候補に挙げら
れた文字を判定候補レジスタに記憶させ、2次判定回路
でこの候補に挙げられた文字のみを詳細に判定するので
、全ての文字パターンを比較するものと比ベコストが低
減する利点を有する。Further, in the present invention, even if a read figure is judged by the primary judgment circuit and the character cannot be identified, the character listed as a candidate is stored in the judgment candidate register, and the secondary judgment circuit is used to identify the character. Since only the listed characters are determined in detail, this method has the advantage of reducing cost compared to a method that compares all character patterns.
第1図は本発明の一実施例による文字認識装置のプロツ
ク図、第2図は文字カテゴリーのパターン図である。
3・・・・・・変換回路、6・・・・・・1次判定回路
、8・・・・・・文字情報レジスタ。FIG. 1 is a block diagram of a character recognition device according to an embodiment of the present invention, and FIG. 2 is a pattern diagram of character categories. 3...Conversion circuit, 6...Primary judgment circuit, 8...Character information register.
Claims (1)
記憶する2値化信号メモリと、この2値化信号メモリの
文字パターンの基準点から文字輪郭をトレースして方向
コード列情報に変換する変換回路と、読取対象の全文字
の標準パターンが記憶されている標準パターンメモリと
、上記変換回路の方向コード列情報および標準パターン
メモリの標準パターンとを所定の線分ごとに比較する1
次判定回路と、この1次判定回路で所定の規準に適合し
ていれば判定候補文字として記憶する判定候補カテゴリ
ーレジスタと、上記1次判定回路で判定した文字の線分
ごとのXY軸座標位置および方向コード列情報を記憶す
る文字特徴レジスタと、上記判定候補カテゴリーレジス
タの判定候補文字ごとに上記文字特徴レジスタに記憶さ
れた信号にもとづき文字判定を行なう2次判定回路とを
具備する文字認識装置。1 A binary signal memory that stores the character pattern obtained by scanning the character to be read and converting it into a binary value, and tracing the character outline from the reference point of the character pattern in this binary signal memory and converting it into direction code string information. A conversion circuit, a standard pattern memory storing standard patterns of all characters to be read, and direction code string information of the conversion circuit and a standard pattern of the standard pattern memory are compared for each predetermined line segment.
a next judgment circuit; a judgment candidate category register that stores characters as judgment candidate characters if they meet predetermined criteria in the first judgment circuit; and an XY-axis coordinate position for each line segment of the character judged by the first judgment circuit. and a character feature register that stores direction code string information, and a secondary determination circuit that performs character determination based on the signal stored in the character feature register for each determination candidate character of the determination candidate category register. .
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP54135613A JPS5941232B2 (en) | 1979-10-19 | 1979-10-19 | character recognition device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP54135613A JPS5941232B2 (en) | 1979-10-19 | 1979-10-19 | character recognition device |
Publications (2)
Publication Number | Publication Date |
---|---|
JPS5659375A JPS5659375A (en) | 1981-05-22 |
JPS5941232B2 true JPS5941232B2 (en) | 1984-10-05 |
Family
ID=15155896
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
JP54135613A Expired JPS5941232B2 (en) | 1979-10-19 | 1979-10-19 | character recognition device |
Country Status (1)
Country | Link |
---|---|
JP (1) | JPS5941232B2 (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS58200379A (en) * | 1982-05-18 | 1983-11-21 | Mitsubishi Electric Corp | Character reader |
-
1979
- 1979-10-19 JP JP54135613A patent/JPS5941232B2/en not_active Expired
Also Published As
Publication number | Publication date |
---|---|
JPS5659375A (en) | 1981-05-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US5113452A (en) | Hand-written character recognition apparatus and method | |
US4319331A (en) | Curve follower | |
US4524454A (en) | Method of assigning direction code to boundary picture element in character recognition system | |
JPS6140684A (en) | Contour tracking device | |
US4975974A (en) | Character recognition apparatus | |
JPS5941232B2 (en) | character recognition device | |
JPH0117191B2 (en) | ||
JP2894111B2 (en) | Comprehensive judgment method of recognition result in optical type character recognition device | |
JPS59158482A (en) | Character recognizing device | |
JPH0418038Y2 (en) | ||
JPH0475557B2 (en) | ||
JPS589471B2 (en) | link link | |
JPS56145472A (en) | Recognizing method of on-line handwritten character | |
JPS59103483A (en) | System for discriminating binary value | |
JPS6055481A (en) | Pattern recognizing device | |
JPS6249482A (en) | Image preprocessing device | |
JPH03229386A (en) | Character recognizing device | |
JPH01217686A (en) | Character reader | |
JPS5814709B2 (en) | Shape recognition method for closed shapes | |
JPH0833924B2 (en) | Drawing reading method | |
JPH0338787A (en) | Character recognition processor | |
JPH0232673B2 (en) | ||
JPS59161784A (en) | On-line character recognizing and rough classifying method | |
JPS6116380A (en) | Area dividing system | |
JPS5935068B2 (en) | Recognition device for characters, figures, etc. |