JPH03271881A - Optical character reader - Google Patents

Abstract

PURPOSE:To improve the recognition rate by removing an abnormal reflected light area from the object to perform comparison and collation at the time of the existence of this area in a scan area and removing a drop-out part from the object to perform coincidence discrimination at the time of the occurrence of partial drop-out in a read input character pattern due to an interposed transparent medium. CONSTITUTION:If the abnormal reflected light area exists, the signal level in an area X of an electric signal is higher than the normal paper surface level and is outputted to a binarizing circuit 20 and an abnormal reflected light area detecting circuit 50, and a binarized picture signal is outputted to a character extracting circuit 30. The character extracting circuit 30 normalizes a character to a fixed size by the inputted picture signal and inputs the result to one input of a collating and discriminating circuit 40. The abnormal reflected light area detecting circuit 50 binarizes the signal based on the paper surface level to detect the abnormal reflected light area and outputs it to an abnormal reflected area position detecting circuit 60 to determine whether the abnormal reflected light area exists or not. The collating and discriminating circuit 40 discriminates the character by pattern matching between standard character patterns and the normalized signal, and the area position is removed from the object of pattern matching. Thus, the recognition rate is improved.

Description

【発明の詳細な説明】 （産業上の利用分野） 本発明は光学式文字読取装置に係り、特に透明媒体を介
して紙表面の文字を読み取る際に生ずる一部欠落文字の
判読技術に関する。DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to an optical character reading device, and more particularly to a technique for reading partially missing characters that occur when reading characters on the surface of paper through a transparent medium.

（従来の技術） 周知のように、光学式文字読取装置は、紙面からの反射
光を電気信号に変換し、その信号レベルの大小関係から
紙面レベルと文字レベルを識別して２値化した文字図形
を形成し、この文字図形と予め用意しである複数種の文
字カテゴリの標準図形とを逐一比較照合して一致度の高
い（つまりは不一致度の低い）標準図形を検出し、その
図形に該当する文字カテゴリをもって入力文字に対する
認識結果の候補として出力するようにしたものである。(Prior Art) As is well known, an optical character reading device converts light reflected from a paper surface into an electrical signal, distinguishes between the paper level and character level based on the magnitude relationship of the signal levels, and reads binary characters. A shape is formed, this character shape is compared point by point with standard shapes of multiple types of character categories prepared in advance, a standard shape with a high degree of matching (that is, a degree of mismatch is low) is detected, and the shape is The corresponding character category is output as a recognition result candidate for the input character.

（発明が解決しようとする課題） 上述した従来の光学式文字読取装置では、透明媒体を介
して紙面の文字を読み取る場合に読取不能あるいは誤読
が生ずるという問題がある。(Problems to be Solved by the Invention) The conventional optical character reading device described above has a problem in that when reading characters on a paper surface through a transparent medium, reading is impossible or misreading occurs.

例えば窓付封書では、窓にセロファン紙やパラフィン紙
等の透明媒体が設けられるが、これにしわやきす等があ
るとそこでの反射光量が増大して対応する文字の一部の
信号レベルが紙面レベルと判断され、その結果読み取っ
た入力文字図形に一部欠落が生じ、標準文字図形との一
致がとれない、あるいは、誤った一致判定を行う場合が
ある。For example, in window envelopes, a transparent medium such as cellophane paper or paraffin paper is provided in the window, but if there are wrinkles or scratches on this, the amount of reflected light increases, and the signal level of some of the corresponding characters will be lower than the surface of the paper. As a result, a part of the read input character figure may be missing, and a match with the standard character figure may not be achieved, or an erroneous match determination may be made.

本発明は、このような問題に鑑みなされたちので、その
目的は、介在する透明媒体での異常反射に起因して、読
み取った入力文字図形に一部欠落が生じた場合に誤読を
減少させ認識率を向上させ得る光学式文字読取装置を提
供することにある。The present invention was developed in view of these problems, and its purpose is to reduce misreading and improve recognition when a part of a read input character figure is missing due to abnormal reflection on an intervening transparent medium. An object of the present invention is to provide an optical character reading device that can improve the reading rate.

（課題を解決するための手段） 前記目的を達成するために、本発明の光学式文字読取装
置は次の如き構成を有する。(Means for Solving the Problems) In order to achieve the above object, the optical character reading device of the present invention has the following configuration.

即ち、本発明の光学式文字読取装置は、光学式文字読取
装置において；　走査変換された電気信号の信号レベル
から走査領域内の異常反射光領域を検出する手段と；　
前記電気信号から形成した入力文字図形と予め用意され
た標準文字図形とを比較照合して文字判定を行う際に前
記異常反射光領域を比較照合の対象領域から除外させる
手段と；　を備えたことを特徴とするものである。That is, in the optical character reading device of the present invention, the optical character reading device includes: means for detecting an abnormally reflected light region within a scanning region from the signal level of a scan-converted electrical signal;
means for excluding the abnormally reflected light area from the area to be compared and verified when character determination is performed by comparing and comparing the input character figure formed from the electric signal with a standard character figure prepared in advance; It is characterized by:

（作　用） 次に、前記の如く構成される本発明の光学式文字読取装
置の作用を説明する。(Function) Next, the function of the optical character reading device of the present invention configured as described above will be explained.

走査領域内に異常反射光領域があるときは、その領域を
除外して比較照合を行う。If there is an abnormally reflected light area within the scanning area, comparison and verification is performed with that area excluded.

その結果、介在する透明媒体に起因して、読み取った入
力文字図形に一部欠落が生じた場合でもその欠落部分を
除外して一致判定を行うので、正しい一致判定が行える
場合が多くなり、誤読が減少し認識率を向上させること
ができる。As a result, even if a part of the read input character figure is missing due to the intervening transparent medium, the missing part is excluded when making a match judgment, which increases the chances of a correct match being made and misreading. can be reduced and the recognition rate can be improved.

（実　施　例） 以下、本発明の実施例を図面を参照して説明する。(Example) Embodiments of the present invention will be described below with reference to the drawings.

第１図は本発明の一実施例に係る光学式文字読取装置を
示す、第１図において、光電変換部１０では、読み取る
べき紙面の所定領域を走査した際の紙面からの反射光を
電気信号に変換する。例えば第２図（ｂ）に示すように
ｒ８５４Ｊなる黒文字の記載された白地用紙の紙面を透
明媒体を介して走査するとし、走査線ＡＢに関し文字「
８コと同「５」の間における文字「５」の近傍領域に対
応する透明媒体にしわやきす等の異常反射光領域がある
場合、走査線ＡＢに係る電気信号の信号レベルは第２図
（ａ）に示すようになり、異常反射光領域に対応する電
気信号の領域Ｘの信号レベルは通常の紙面レベルよりも
２〜３倍大きなレベルとなる。このような信号が２値化
回路２０と異常反射光領域検出回路５０とへ出力される
。FIG. 1 shows an optical character reading device according to an embodiment of the present invention. In FIG. 1, a photoelectric conversion unit 10 converts light reflected from the paper surface when scanning a predetermined area of the paper surface to be read into an electrical signal. Convert to For example, as shown in FIG. 2(b), suppose that the surface of a blank sheet of paper with black letters r854J written on it is scanned through a transparent medium.
If there is an abnormal reflected light area such as wrinkles or scratches on the transparent medium corresponding to the area near the character ``5'' between 8 and ``5'', the signal level of the electrical signal related to scanning line AB will be as shown in Figure 2. As shown in (a), the signal level in region X of the electric signal corresponding to the abnormal reflected light region is two to three times higher than the normal level on the page. Such a signal is output to the binarization circuit 20 and the abnormal reflected light area detection circuit 50.

２値化回路２０では、入力された電気信号について紙面
レベルと文字レベルを区別する２値化処理を行い、その
２値化処理した画像信号を文字抽出回路３０へ出力する
。第３図は原画像（第２図（ｂ））の全体についての電
気信号を２値化して得られた図形を示し、文字「５」の
点線で囲まれた領域（イ）は前記異常反射光によって欠
落した部分を示す。The binarization circuit 20 performs a binarization process on the input electrical signal to distinguish between the paper level and the character level, and outputs the binarized image signal to the character extraction circuit 30 . Figure 3 shows a figure obtained by binarizing the electric signal for the entire original image (Figure 2 (b)), and the area (a) surrounded by the dotted line of the letter "5" is the abnormal reflection. Light shows the missing parts.

文字抽出回路３０では、入力された画像信号から例えば
公知の技術である投影法によって文字を切り出して１文
字毎に分離するとともに、その１文字を一定の大きさに
正規化する。例えば第４図は文字「５」の正規化パター
ンを示すが、縦が０〜Ｆの１６メツシユ、横が０〜９の
１０メツシユに正規化され、各メツシュには文字信号部
（黒レベル）であれば論理°゛１”　（図中Ｘ印で示す
）が設定され、紙面部（白レベル）であれば論理゛Ｏ”
（図中ブランクで示す）が設定される。なお、第４図中
の黒枠部（ロ）は前記欠落領域（イ）を示し白レベルと
なっている。このような正規化信号が照合判定回路４０
の一方の入力となる。The character extraction circuit 30 cuts out characters from the input image signal using, for example, a well-known projection method, separates each character, and normalizes each character to a certain size. For example, Figure 4 shows the normalized pattern for the character "5", but it is normalized into 16 meshes from 0 to F in the vertical direction and 10 meshes from 0 to 9 in the horizontal direction, and each mesh has a character signal part (black level). If so, logic °゛1” (indicated by
(shown as blank in the figure) is set. Note that the black frame portion (b) in FIG. 4 indicates the missing area (a) and has a white level. Such a normalized signal is used by the matching determination circuit 40.
This is one of the inputs.

一方、異常反射光領域検出回路５０では、入力された電
気信号について紙面レベルを基準に２値化して異常反射
光領域を検出しく第５図）、その検出信号を異常反射光
領域位置検出回路６０へ出力する。On the other hand, the abnormal reflected light area detection circuit 50 binarizes the input electric signal based on the page level to detect the abnormal reflected light area (FIG. 5), and sends the detection signal to the abnormal reflected light area position detection circuit 60. Output to.

異常反射光領域位置検出回路６０では、文字抽出回路３
０において作成される正規化パターンのいずれの位置に
異常反射光領域が在るかを入力された検出信号に基づき
決定する。即ち、第４図中の黒枠部（ロ）の座標位置を
検出するのである。In the abnormal reflected light area position detection circuit 60, the character extraction circuit 3
Based on the input detection signal, it is determined at which position of the normalized pattern created at 0 the abnormal reflected light area is located. That is, the coordinate position of the black frame (b) in FIG. 4 is detected.

この位置検出信号は照合判定回路４０の他方の入力とな
る。This position detection signal becomes the other input of the matching determination circuit 40.

照合判定回路４０は、基本的には、辞書７０に予め格納
しである標準文字図形と文字抽出回路３０からの正規化
信号とのパターンマツチングを行い文字を判定する。そ
の際に、位置検出信号が正規化信号に対するゲート信号
として作用し、正規化パターンにおいて位置検出信号が
指定した領域位置はパターンマツチングの対象から除外
されるようになっている。以下、文字「５」についての
判定照合を例に挙げて具体的に説明する。なお、辞書７
０に格納される文字カテゴリ「５」の標準文字図形は、
例えば第８図に示すように、紙面部（白レベル）である
べき領域を論理“１”　（［×」印）で表した標準文字
図形（第８図（ａ））と、文字信号部（黒レベル）であ
るべき領域を論理“１”（「×ｊ印）で表した標準文字
図形（第８図（ｂ））とからなる、この標準文字図形の
大きさは、前記正規化パターンと同様に１６Ｘ１０メツ
シユで構成されることは勿論である。The match determination circuit 40 basically performs pattern matching between standard character figures stored in advance in the dictionary 70 and the normalized signal from the character extraction circuit 30 to determine characters. At this time, the position detection signal acts as a gate signal for the normalization signal, and the area position specified by the position detection signal in the normalization pattern is excluded from the pattern matching target. Hereinafter, a detailed explanation will be given by taking as an example the determination collation for the character "5". In addition, dictionary 7
The standard character figure of character category "5" stored in 0 is:
For example, as shown in Fig. 8, a standard character figure (Fig. 8 (a)) in which the area that should be the paper surface area (white level) is represented by a logic "1" ([x] mark), and a character signal part ( The size of this standard character figure (Fig. 8(b)) is made up of a standard character figure (Fig. 8(b)) in which the area where the black level should be is expressed as a logical "1"("xjmark"). Of course, it is also composed of a 16×10 mesh.

第６図は照合判定回路４０の具体的構成例を示す、第６
図において、文字抽出回路３０の出力たる正規化信号は
反転回路４２０と比較回路４２１とに入力する。また、
異常反射光領域位置検出口８６０の出力たる位置信号は
反転回路４２０に入力する。FIG. 6 shows a specific configuration example of the collation determination circuit 40.
In the figure, the normalized signal output from the character extraction circuit 30 is input to an inversion circuit 420 and a comparison circuit 421. Also,
The position signal output from the abnormal reflected light area position detection port 860 is input to the inversion circuit 420.

そして、辞書７０からの標準文字図形（白レベル）は比
較回路４２１に入力し、標準文字図形（゛黒レベル）は
比較回路４２２に入力する。The standard character figures (white level) from the dictionary 70 are input to a comparison circuit 421, and the standard character figures (black level) are input to a comparison circuit 422.

まず、反転回路４２０では、正規化信号の各メ・ンシュ
の論理状態を反転する。その際に位置信号はゲート信号
として機能し、正規化パターン（第４図）における欠落
領域（口〉の論理状態は変更されずそのまま論理“Ｏ”
　（ブランク）の状態が保持される（第７図）。First, the inverting circuit 420 inverts the logic state of each mesh of the normalized signal. At that time, the position signal functions as a gate signal, and the logic state of the missing area (mouth) in the normalized pattern (Figure 4) remains unchanged and becomes logic "O".
(blank) state is maintained (FIG. 7).

比較回路４２１と同４２２は、両人力の各メツシュごと
の論理積をとることによって一致不一致を判定し、一致
しなければ不一致信号を不一致数処理回路４２３へ出力
する０例えば、比較回路４２１では、正規化パターン（
第４図）と標準文字図形（第８図（ａ））との比較をす
るが、横のＯ〜９の列における第４列と縦の０〜Ｆの行
における第０行との交点位置のメツシュでは、論理値が
共に“１”であるから、論理積は１”となり、不一致信
号が出力される。しかし、欠落領域（ロ）の比較では論
理積は０”となるから不一致信号は出力されない、また
、比較回路４２２では、反転正規化パターン（第７図〉
と標準文字図形（第８図（ｂ））との比較をするが、第
４列と第Ｆ行との交点位置のメツシュでは論理値が共に
“１″であるから、論理積は“１”となり不一致信号が
出力される。しかし、欠落領域（ロ）の比較では、論理
積は°０”となるから不一致信号は出力されない０以上
要するに、比較回路４２１と同４２２では、異常反射光
領域を除外して各標準文字図形ごとに一致不一致の判定
照合をしているのである。Comparison circuits 421 and 422 determine whether or not they match by performing a logical product for each mesh of both people's power, and if they do not match, output a mismatch signal to mismatch number processing circuit 423. For example, in comparison circuit 421, Normalization pattern (
Figure 4) and the standard character figure (Figure 8 (a)) are compared, and the intersection position of the 4th column in the horizontal columns O to 9 and the 0th row in the vertical rows 0 to F. In the mesh, both logical values are "1", so the logical product is 1, and a mismatch signal is output.However, when comparing the missing area (b), the logical product is 0, so the mismatch signal is In addition, the comparator circuit 422 outputs an inverted normalization pattern (see FIG. 7).
and the standard character figure (Figure 8 (b)), but since the logical values at the mesh at the intersection of the 4th column and the F row are both "1", the logical product is "1". Therefore, a mismatch signal is output. However, in the comparison of missing areas (b), the logical product is 0'', so no mismatch signal is output. It is used to check whether there is a match or a mismatch.

不一致数処理回路４２３では、両比較回路（４２１゜４
２２）が出力する不一致信号を各標準文字図形ごとに計
数し、その計数結果を判定回路４２４に出力する。In the mismatch number processing circuit 423, both comparison circuits (421°4
22) is counted for each standard character figure, and outputs the counting result to the determination circuit 424.

判定回路４２４では、不一致数の最も少ない文字を判定
し、それを認識結果の第１候補として出力する。The determination circuit 424 determines the character with the least number of mismatches and outputs it as the first candidate of the recognition result.

（発明の効果） 以上説明したように、本発明の光学式文字読取装置によ
れば、走査領域内に異常反射光領域があるときは、その
領域を除外して比較照合を行うようにしたので、介在す
る透明媒体に起因して、読み取った入力文字図形に一部
欠落が生じた場合でもその欠落部分を除外して一致判定
を行うことができ、正しい一致判定が行える場合が多く
なり、誤読が減少し認識率を向上させることができる効
果がある。(Effects of the Invention) As explained above, according to the optical character reading device of the present invention, when there is an abnormally reflected light area within the scanning area, comparison and verification is performed by excluding that area. , even if a part of the read input character figure is missing due to the intervening transparent medium, the missing part can be excluded when making a match judgment, which increases the chances of correct match judgment and prevents misreading. This has the effect of reducing the number of errors and improving the recognition rate.

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

第１図は本発明の一実施例に係る光学式文字読取装置の
構成ブロック図、第２図は光電変換部の動作説明図、第
３図は２値化回路の動作説明図、第４図は文字抽出回路
が作成する正規化パターンの一例を示す図、第５図は異
常反射光領域検出回路の動作説明図、第６図は照合判定
回路の具体例を示す構成ブロック図、第７図は反転回路
が出力する反転正規化パターンの一例を示す図、第８図
は辞書に格納される標準文字図形例を示す図である。 １０・・・・・・光電変換部、　２０・・・・・・２値
化回路、３０・・・・・・文字抽出回路、　４０・・・
・・・照合判定回路、５０・・・・・・異常反射光領域
検出回路、　６０・・・・・・異常反射光領域位置検出
回路、 ７０・・・・・・辞書。FIG. 1 is a block diagram of the configuration of an optical character reading device according to an embodiment of the present invention, FIG. 2 is an explanatory diagram of the operation of the photoelectric conversion section, FIG. 3 is an explanatory diagram of the operation of the binarization circuit, and FIG. 5 is a diagram illustrating an example of a normalized pattern created by the character extraction circuit, FIG. 5 is an explanatory diagram of the operation of the abnormal reflected light area detection circuit, FIG. 6 is a configuration block diagram illustrating a specific example of the matching determination circuit, and FIG. 7 8 is a diagram showing an example of an inverted normalization pattern outputted by an inverting circuit, and FIG. 8 is a diagram showing an example of standard character figures stored in a dictionary. 10...Photoelectric conversion unit, 20...Binarization circuit, 30...Character extraction circuit, 40...
...Verification determination circuit, 50... Abnormal reflected light area detection circuit, 60... Abnormal reflected light area position detection circuit, 70... Dictionary.

Claims (1)

【特許請求の範囲】[Claims] 光学式文字読取装置において；走査変換された電気信号
の信号レベルから走査領域内の異常反射光領域を検出す
る手段と；前記電気信号から形成した入力文字図形と予
め用意された標準文字図形とを比較照合して文字判定を
行う際に前記異常反射光領域を比較照合の対象領域から
除外させる手段と；を備えたことを特徴とする光学式文
字読取装置。In an optical character reading device; means for detecting an abnormally reflected light area within a scanning area from the signal level of a scan-converted electric signal; and an input character figure formed from the electric signal and a standard character figure prepared in advance. An optical character reading device comprising: means for excluding the abnormal reflected light area from a target area for comparison and verification when character determination is performed by comparison and verification.