JPH04316176A - Name card recognizing method and name card managing machine - Google Patents

Name card recognizing method and name card managing machine

Info

Publication number
JPH04316176A
JPH04316176A JP3083754A JP8375491A JPH04316176A JP H04316176 A JPH04316176 A JP H04316176A JP 3083754 A JP3083754 A JP 3083754A JP 8375491 A JP8375491 A JP 8375491A JP H04316176 A JPH04316176 A JP H04316176A
Authority
JP
Japan
Prior art keywords
business card
name card
vertical
character
characters
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
JP3083754A
Other languages
Japanese (ja)
Inventor
Takashi Nitta
隆志 新田
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Seiko Epson Corp
Original Assignee
Seiko Epson 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 Seiko Epson Corp filed Critical Seiko Epson Corp
Priority to JP3083754A priority Critical patent/JPH04316176A/en
Publication of JPH04316176A publication Critical patent/JPH04316176A/en
Pending legal-status Critical Current

Links

Abstract

PURPOSE:To automatically and accurately input name card data at a high speed by discriminating whether a name card is of a vertical or horizontal type from the aspect ratio of the circumscribed rectangle of each character and discriminating the items of the name card by utilizing the discriminated result of the aspect ratio. CONSTITUTION:A picture reader 401 reads a name card and a picture memory 402 stores the read binary data. A CPU 403 performs an 8-coupling component labeling process on the binarized picture of the name card and finds the circumscribed rectangle of the same label element by giving the same label to coupled picture elements. The CPU 403 discriminates whether the name card is of a vertical or horizontal type by checking the aspect ratio of the circumscribed rectangle of each character. Then the CPU 403 segments individual characters by using the vertical or horizontal type information, recognizes the segmented characters, and converts the recognized characters into character codes. Thereafter, the CPU 403 discriminates each character string which represents a company name, personal name, address, telephone number, etc., from the recognized result by performing a knowledge process, etc. The name card data obtained as a result of each process are stored in a memory 406 where the name card data are stored as electronic data.

Description

【発明の詳細な説明】[Detailed description of the invention]

【0001】0001

【産業上の利用分野】本発明は名刺認識方法および名刺
管理機に関する。
BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a business card recognition method and a business card management machine.

【0002】0002

【従来の技術】名刺はその書式に大きく分けて縦型と横
型が存在し、互いに文字の向きが違う。
BACKGROUND OF THE INVENTION Business cards can be roughly divided into vertical and horizontal formats, and the orientation of the characters is different.

【0003】従来この様に文字の向きが異なる名刺を認
識する場合には、特開昭62−166479のように最
初に名刺画像から文字パターンを切り出し、その文字パ
ターンとあらかじめ用意した0゜,90゜,180゜,
270゜回転した、1文字あたり4つの比較パターンと
を比較、認識して、一番一致度が高い比較パターンの向
きをその名刺の向きとするというものであった。
Conventionally, when recognizing business cards with different character orientations, as in Japanese Patent Laid-Open No. 62-166479, a character pattern is first cut out from a business card image, and the character pattern and a pre-prepared 0°, 90°゜,180゜、
The system compares and recognizes four comparison patterns per character rotated by 270 degrees, and sets the orientation of the comparison pattern with the highest degree of matching as the orientation of the business card.

【0004】0004

【発明が解決しようとする課題】しかし、前述の従来技
術では 1)一度、文字を認識することになり、しかもひとつの
文字に対して4回文字認識することになるため処理時間
が増大する。
However, in the prior art described above, 1) a character is recognized once, and each character is recognized four times, which increases the processing time.

【0005】2)ひとつの文字に対して4つの比較用パ
ターンを持たなければならず、そのための記憶容量が大
きくなる。
2) It is necessary to have four comparison patterns for one character, which increases the storage capacity.

【0006】3)文字の向きが決定されない前に、一度
文字を切り出さねばならないために、レイアウトが複雑
であったり、文字の大きさがばらばらな名刺の場合、間
違った文字パターンを切り出す可能性が高く、そのため
間違った文字の向きを判断する可能性がある。
3) Since the characters must be cut out before the orientation of the characters is determined, there is a possibility that the wrong character pattern will be cut out in the case of business cards with complex layouts or variable font sizes. high, which may result in incorrect character orientation being determined.

【0007】そこで、本発明はこのような問題点を解決
するもので、その目的とするところは、非常に高速に文
字の向きの判別が可能で、記憶容量が少なくてよく、レ
イアウトが複雑な場合でも正確に文字の向きの検出が可
能な名刺認識方法、および複数枚の名刺データをオペレ
ータの指示を必要とせずに自動、高速、正確に入力可能
で、かつ安価な名刺管理機を提供することにある。
[0007]The present invention is intended to solve these problems.The purpose of the present invention is to be able to determine the orientation of characters very quickly, to require less memory capacity, and to be able to solve problems with complex layouts. To provide a business card recognition method capable of accurately detecting the orientation of characters even in the case of a business card, and an inexpensive business card management machine capable of automatically, quickly and accurately inputting data on multiple business cards without requiring operator instructions. There is a particular thing.

【0008】[0008]

【課題を解決するための手段】本発明の名刺認識方法は
、レイアウト自由な名刺を、画像読み取り装置によって
読み取り、その読み取った名刺画像から、個々の文字の
外接矩形の縦横比によって名刺が縦型か横型かを自動で
判別し、その判別結果を利用して文字を切り出し、文字
を認識し、名刺項目を判別することを特徴とする。
[Means for Solving the Problems] The business card recognition method of the present invention reads a business card with a free layout using an image reading device, and from the read business card image, the business card is vertically shaped according to the aspect ratio of the circumscribed rectangle of each character. It is characterized by automatically determining whether the card is horizontal or horizontal, cutting out characters using the determination results, recognizing the characters, and determining business card items.

【0009】また、本発明の名刺管理機は、上述の名刺
認識方法をデータ入力部に用いることにより、レイアウ
ト自由な複数枚の名刺データを自動で高速に入力するこ
とを特徴とする。
Further, the business card management machine of the present invention is characterized in that by using the above-described business card recognition method in the data input section, data for a plurality of business cards with a free layout can be automatically and quickly input.

【0010】0010

【実施例】【Example】

(実施例1)以下、本発明を実施例に基づいて詳細に説
明する。
(Example 1) Hereinafter, the present invention will be explained in detail based on an example.

【0011】図1は本発明の名刺認識方法における名刺
の縦型、横型判別方法の説明図である。図2は本発明の
名刺認識方法のフローチャートである。図3は本発明の
名刺認識方法における名刺の縦型、横型判別方法のフロ
ーチャートである。図4は本発明の名刺管理機の構成図
である。
FIG. 1 is an explanatory diagram of a method for determining whether a business card is vertical or horizontal in the business card recognition method of the present invention. FIG. 2 is a flowchart of the business card recognition method of the present invention. FIG. 3 is a flowchart of a method for determining whether a business card is vertical or horizontal in the business card recognition method of the present invention. FIG. 4 is a block diagram of the business card management machine of the present invention.

【0012】以下、図面を用いて本発明の名刺認識方法
を詳細に説明する。
The business card recognition method of the present invention will be explained in detail below with reference to the drawings.

【0013】まず、図2は本発明の名刺認識方法のフロ
ーチャートである。画像読み取り装置で読み取られ、既
定のしきい値で2値化された名刺画像に対して8連結成
分のラベリング処理(または8連結成分の境界追跡処理
)を行ない、連結している画素に同じラベルを与える(
1)。(1)のラベリング結果に対して、同一ラベル要
素の外接矩形を求める(2)。これ以後はこの外接矩形
の情報だけを用いて、文字の切り出し、縦型、横型の判
別をすることにより、画像をそのまま用いる場合に比べ
て処理の簡略化、高速化がはかれる。つぎに、文字列を
切り出す(3)。名刺の場合文字列どうしの間隔が比較
的大きいために切り出しは容易であり、その結果その文
字列に含まれる文字の大きさの推定が可能となる。つぎ
に、(2)で求めた外接矩形、および(3)で求めた文
字列およびその文字列に含まれる文字の大きさの情報を
用いて名刺の縦型、横型を判別する(4)。つぎに、(
4)で求めた縦型、横型の情報を用いて(3)の文字列
から個々の文字を切り出す(5)。つぎに、(5)で切
り出された文字を認識し文字コード化する(6)。 (6)の認識結果に対してどの文字列が名刺の会社名、
氏名、住所、電話番号などの、どの項目に該当するかを
知識処理などによって判別し(7)、名刺認識が終了す
る。もし認識結果が思わしくない場合には順次前に戻っ
て処理を繰り返す。
First, FIG. 2 is a flowchart of the business card recognition method of the present invention. 8-connected component labeling processing (or 8-connected component boundary tracking processing) is performed on the business card image that has been read by an image reading device and binarized using a predetermined threshold value, and connected pixels are labeled with the same label. give(
1). For the labeling result in (1), a circumscribed rectangle of the same label element is determined (2). From now on, only the information on this circumscribed rectangle is used to cut out characters and determine whether they are vertical or horizontal, thereby simplifying and speeding up the processing compared to the case where the image is used as is. Next, cut out the character string (3). In the case of a business card, the intervals between character strings are relatively large, so cutting out is easy, and as a result, it is possible to estimate the size of the characters included in the character string. Next, the vertical or horizontal type of the business card is determined using the circumscribed rectangle obtained in (2), the character string obtained in (3), and information on the size of the characters contained in the character string (4). next,(
Using the vertical and horizontal information obtained in 4), cut out individual characters from the character string in (3) (5). Next, the characters extracted in (5) are recognized and converted into character codes (6). Based on the recognition result of (6), which character string is the company name of the business card,
Knowledge processing or the like is used to determine which item, such as name, address, telephone number, etc., applies (7), and business card recognition is completed. If the recognition result is unsatisfactory, go back and repeat the process.

【0014】図1は名刺の縦型、横型判別方法の説明図
である。図1(a)は横型の名刺、図1(b)は縦型の
名刺の例である。図1を見てわかるように名刺を図1の
ように置いた場合、図1(a)の横型の場合は文字は縦
向きに、図1(b)の縦型の場合は文字は横向きに書か
れていることになる。そのため、それぞれの文字の外接
矩形を見ると、横型は101のような縦長の外接矩形、
縦型は102のような横長の外接矩形が多くなる。この
ことを利用して文字の外接矩形の縦横比を調べることに
よって名刺が縦型か横型かの判別が可能となる。図3は
実際の名刺の縦型、横型判別処理のフローチャートであ
る。外接矩形の大きさがその外接矩形が含まれる文字列
の文字の大きさに比較してある一定値以下である場合、
その外接矩形は判別に使用しないことにする(8)。つ
ぎに、外接矩形の縦横比を求める(9)。つぎに、縦横
比が極端に大きいものや、小さいものは1や一など判別
に利用できないものなので、使用しないことにする(1
0)。つぎに、縦長な外接矩形の数とその縦横比の平均
値を求める(11)。横長な外接矩形の数とその縦横比
の平均値を求める(12)。全ての外接矩形の縦横比の
平均値を求める(13)。(11)、(12)、(13
)で求めた値を用いて名刺が縦型か、横型かを判定する
(14)。判定は(11)、(12)、(13)で求め
た値それぞれの関係を用いて決定されるものであるが、
一例としては(11)で求めた縦長の外接矩形の数が、
(12)で求めた横長の外接矩形の数よりもある一定数
以上多く、かつ(13)で求めた全ての外接矩形の縦横
比の平均値が縦長の矩形を表わすものであった場合、名
刺は横型であると判定するものである。
FIG. 1 is an explanatory diagram of a method for determining whether a business card is vertical or horizontal. FIG. 1(a) is an example of a horizontal business card, and FIG. 1(b) is an example of a vertical business card. As you can see from Figure 1, when a business card is placed as shown in Figure 1, the characters will be oriented vertically in the horizontal format shown in Figure 1 (a), and the characters will be horizontal in the vertical format shown in Figure 1 (b). It will be written. Therefore, if you look at the circumscribed rectangle of each character, the horizontal type is a vertically long circumscribed rectangle like 101,
The vertical type has many horizontally long circumscribed rectangles like 102. By utilizing this fact and checking the aspect ratio of the circumscribed rectangle of the characters, it is possible to determine whether the business card is vertical or horizontal. FIG. 3 is a flowchart of an actual business card vertical/horizontal type determination process. If the size of the circumscribed rectangle is less than a certain value compared to the size of the characters in the string that includes the circumscribed rectangle,
The circumscribed rectangle is not used for determination (8). Next, the aspect ratio of the circumscribed rectangle is determined (9). Next, we will not use items with extremely large or small aspect ratios, as they cannot be used to distinguish between 1 and 1.
0). Next, the number of vertically long circumscribed rectangles and the average value of their aspect ratios are determined (11). The number of horizontally long circumscribed rectangles and the average value of their aspect ratios are calculated (12). The average value of the aspect ratios of all circumscribed rectangles is determined (13). (11), (12), (13
) is used to determine whether the business card is vertical or horizontal (14). The determination is made using the relationship between the values obtained in (11), (12), and (13), but
As an example, the number of vertically long circumscribed rectangles found in (11) is
If the number of horizontally long circumscribed rectangles is greater than the number of horizontally long circumscribed rectangles found in (12), and the average aspect ratio of all the circumscribed rectangles found in (13) represents a vertically long rectangle, then the business card is determined to be horizontal.

【0015】つぎに、本発明の名刺認識装置を用いた名
刺管理機の構成図を図4に示す。名刺は画像読み取り装
置401で読み取られ、既定のしきい値で2値化された
2値データが画像用メモリー402に記憶される。その
2値化画像を用いてCPU403で演算し、8連結ラベ
リング、外接矩形取り出し、文字列切り出し、縦型、横
型判別、文字切り出し、文字認識、項判別の処理をする
。各処理の結果求められた名刺データは名刺データ保存
用メモリー406に記憶され、データの検索、表示など
必要なときに随時取り出される。データの表示は表示装
置404によって行なわれる。オペレーターによって与
えられる指示は指示入力装置405によってCPU40
3へ与えられる。ワークRAM407はCPU403が
処理を行なうための作業領域として使用される。ここで
、本発明の名刺認識方法を使用することにより名刺の縦
型、横型を自動で高速に判別することが可能なため、画
像読み取り装置401に名刺を連続で読み取る機能を持
たせることによって、複数枚の名刺データの入力がオペ
レータの指示を必要とせずに、自動で高速に行なうこと
が可能となり、しかも1枚の名刺を認識している間に次
の名刺画像を読み取ることによってさらなる高速化が可
能となる。
Next, FIG. 4 shows a block diagram of a business card management machine using the business card recognition device of the present invention. A business card is read by an image reading device 401, and binary data that is binarized using a predetermined threshold value is stored in an image memory 402. Using the binarized image, the CPU 403 performs calculations such as 8-concatenation labeling, circumscribing rectangle extraction, character string extraction, vertical/horizontal type discrimination, character extraction, character recognition, and term discrimination. The business card data obtained as a result of each process is stored in a business card data storage memory 406, and retrieved whenever necessary for data retrieval, display, etc. Display of data is performed by display device 404. Instructions given by the operator are sent to the CPU 40 by the instruction input device 405.
given to 3. The work RAM 407 is used as a work area for the CPU 403 to perform processing. Here, by using the business card recognition method of the present invention, it is possible to automatically and quickly determine whether the business card is vertical or horizontal, so by providing the image reading device 401 with a function to read business cards continuously, It is now possible to input data for multiple business cards automatically and quickly without requiring operator instructions, and even faster by reading the next business card image while one business card is being recognized. becomes possible.

【0016】以上のように、本発明の名刺認識方法は、
名刺の文字の外接矩形を求め、文字列を切り出して文字
の大きさを推定する。そして、その外接矩形の縦横比を
求め、その縦横比と文字の大きさの情報を使用して名刺
が縦型か横型かを判別する。その判別結果を利用して文
字を切り出す。その切り出された文字に対して文字認識
し、文字コード化する。その認識結果に対して知識処理
などを適用することにより会社名や、氏名、住所等の項
目を判別するというものである。このため、名刺データ
を入力する場合、縦型、横型の違いをオペレータが指示
する必要がない。また、名刺の縦型、横型の判別に認識
などの複雑な処理をする必要がないため非常に高速に判
別することが可能である。また、縦型、横型判別用の認
識パターン辞書を持たなくてよいために記憶容量も少な
くてよい。また、文字の切り出しなど、縦型、横型の事
前情報を必要とする処理を行なわないために、レイアウ
トが複雑で文字の大きさがばらばらな名刺に対しても正
確な縦型、横型判別が可能で、結果として正確な名刺認
識が可能となる。
As described above, the business card recognition method of the present invention
Find the circumscribed rectangle of the characters on the business card, cut out the character string, and estimate the size of the characters. Then, the aspect ratio of the circumscribed rectangle is determined, and the aspect ratio and character size information are used to determine whether the business card is vertical or horizontal. The characters are cut out using the discrimination results. The extracted characters are recognized and converted into character codes. By applying knowledge processing to the recognition results, items such as company name, name, address, etc. are determined. Therefore, when inputting business card data, there is no need for the operator to indicate the difference between vertical and horizontal formats. Further, since there is no need to perform complicated processing such as recognition to determine whether a business card is vertical or horizontal, it is possible to perform the determination very quickly. Furthermore, since there is no need to have a recognition pattern dictionary for determining vertical and horizontal formats, the storage capacity may be small. In addition, since there is no processing that requires prior information about vertical and horizontal orientation, such as cutting out characters, it is possible to accurately determine vertical and horizontal orientation even for business cards with complex layouts and varying font sizes. As a result, accurate business card recognition becomes possible.

【0017】よって、本発明によるとレイアウト自由な
名刺のデータを、オペレータの指示を必要とせずに自動
で、高速に入力可能な名刺認識方法を提供可能となる。 また、この名刺認識方法を名刺データ入力部に用いて、
画像読み取り装置に名刺を連続で読み取る機能を持たせ
ることにより、複数枚の名刺データをオペレータの指示
を必要とせずに、自動で高速に入力可能で、しかも1枚
の名刺を認識している間に次の名刺画像の読み取りを行
なうことによってさらなる高速化が可能な、名刺のレイ
アウトを限定しなくてもよい、安価な名刺管理機を提供
可能となる。
Therefore, according to the present invention, it is possible to provide a business card recognition method that allows business card data with a free layout to be input automatically and at high speed without requiring instructions from an operator. Also, by using this business card recognition method in the business card data input section,
By equipping the image reading device with the ability to read business cards continuously, data on multiple business cards can be entered automatically and at high speed without the need for operator instructions, while still recognizing a single business card. It is possible to provide an inexpensive business card management machine that can further increase the speed by reading the next business card image at the same time, and does not require restrictions on the layout of business cards.

【0018】また、本発明の名刺認識方法は一般の文書
を認識する場合にも利用可能なものである。
The business card recognition method of the present invention can also be used to recognize general documents.

【0019】[0019]

【発明の効果】以上のように本発明の名刺認識方法は名
刺の縦型、横型を文字の外接矩形の縦横比の情報だけを
用いて判別するものであり、オペレータの指示を必要と
せずに自動で名刺認識が可能である。また、縦型、横型
の判別に文字認識などの複雑な処理を必要としないため
高速である。また、判別用の認識パターン辞書が必要無
いため記憶容量も少なくてよい。また、文字切り出しの
ように縦型、横型の事前情報を必要とする処理ではない
ために、レイアウトが複雑で文字の大きさがばらばらな
名刺に対しても正確な縦型、横型の判別が可能で、結果
として正確な名刺認識が可能となる。
[Effects of the Invention] As described above, the business card recognition method of the present invention determines whether a business card is vertical or horizontal using only the information on the aspect ratio of the circumscribed rectangle of the characters, without requiring operator instructions. Automatic business card recognition is possible. In addition, it is fast because it does not require complex processing such as character recognition to distinguish between vertical and horizontal formats. Furthermore, since there is no need for a recognition pattern dictionary for discrimination, the storage capacity may be small. In addition, unlike text extraction, which does not require prior information about vertical and horizontal orientation, it is possible to accurately determine vertical and horizontal orientation even for business cards with complex layouts and varying font sizes. As a result, accurate business card recognition becomes possible.

【0020】また、本発明の名刺認識方法を名刺データ
入力部に用いた名刺管理機は、画像読み取り装置に名刺
を連続で読み取る機能を持たせることにより、レイアウ
トが自由な複数枚の名刺データをオペレータの指示を必
要とせずに、自動、高速、正確に入力可能で、装置構成
が簡単で安価なものとなる。
[0020] Furthermore, a business card management machine using the business card recognition method of the present invention in the business card data input section has an image reading device with a function of reading business cards in succession, so that data on a plurality of business cards with a free layout can be processed. Automatic, high-speed, and accurate input is possible without requiring operator instructions, and the device configuration is simple and inexpensive.

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

【図1】本発明の名刺認識方法における名刺の縦型、横
型判別方法の説明図。
FIG. 1 is an explanatory diagram of a method for determining whether a business card is vertical or horizontal in the business card recognition method of the present invention.

【図2】本発明の名刺認識方法のフローチャート。FIG. 2 is a flowchart of the business card recognition method of the present invention.

【図3】本発明の名刺認識方法における名刺の縦型、横
型判別方法のフローチャート。
FIG. 3 is a flowchart of a method for determining whether a business card is vertical or horizontal in the business card recognition method of the present invention.

【図4】本発明の名刺管理機の構成図。FIG. 4 is a configuration diagram of a business card management machine of the present invention.

【符号の説明】[Explanation of symbols]

101  横型名刺における縦長の外接矩形102  
縦型名刺における横長の外接矩形401  画像読み取
り装置
101 Vertical circumscribed rectangle 102 in a horizontal business card
Horizontally long circumscribed rectangle 401 in a vertical business card Image reading device

Claims (2)

【特許請求の範囲】[Claims] 【請求項1】  レイアウト自由な名刺を、画像読み取
り装置によって読み取り、その読み取った名刺画像から
、個々の文字の外接矩形の縦横比によって名刺が縦型か
横型かを自動で判別し、その判別結果を利用して文字を
切り出し、文字を認識し、名刺項目を判別することを特
徴とする名刺認識方法。
[Claim 1] A business card with a free layout is read by an image reading device, and based on the read business card image, it is automatically determined whether the business card is vertical or horizontal based on the aspect ratio of the circumscribed rectangle of each character, and the determination result is A business card recognition method characterized by cutting out characters using , recognizing the characters, and determining business card items.
【請求項2】  少なくとも請求項1記載の名刺認識方
法をデータ入力部に用いることにより、レイアウト自由
な複数枚の名刺データを自動で高速に入力することを特
徴とする名刺管理機。
2. A business card management machine, characterized in that by using at least the business card recognition method according to claim 1 in a data input section, data for a plurality of business cards with a free layout can be automatically input at high speed.
JP3083754A 1991-04-16 1991-04-16 Name card recognizing method and name card managing machine Pending JPH04316176A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP3083754A JPH04316176A (en) 1991-04-16 1991-04-16 Name card recognizing method and name card managing machine

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP3083754A JPH04316176A (en) 1991-04-16 1991-04-16 Name card recognizing method and name card managing machine

Publications (1)

Publication Number Publication Date
JPH04316176A true JPH04316176A (en) 1992-11-06

Family

ID=13811335

Family Applications (1)

Application Number Title Priority Date Filing Date
JP3083754A Pending JPH04316176A (en) 1991-04-16 1991-04-16 Name card recognizing method and name card managing machine

Country Status (1)

Country Link
JP (1) JPH04316176A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7711642B2 (en) 2000-05-31 2010-05-04 Fujitsu Limited Payment form discrimination method and apparatus

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7711642B2 (en) 2000-05-31 2010-05-04 Fujitsu Limited Payment form discrimination method and apparatus

Similar Documents

Publication Publication Date Title
US6038351A (en) Apparatus and method for multi-entity, mixed document environment document identification and processing
Yuan et al. Text extraction from gray scale document images using edge information
CN112800848A (en) Structured extraction method, device and equipment of information after bill identification
JPH0430070B2 (en)
JPH04316176A (en) Name card recognizing method and name card managing machine
US7133556B1 (en) Character recognition device and method for detecting erroneously read characters, and computer readable medium to implement character recognition
JPH0247788B2 (en)
CN115131806B (en) Method and system for identifying OCR (optical character recognition) image information of various certificates based on deep learning
CN114202761B (en) Information batch extraction method based on picture information clustering
JPH10302025A (en) Handwritten character recognizing device and its program recording medium
JP3220226B2 (en) Character string direction determination method
KR100234028B1 (en) Character area discriminating method of character recognizing system
JP2001291058A (en) Character recognizing device and recording medium
JP2972443B2 (en) Character recognition device
JPH04260980A (en) Device for recognizing graphic
Thi et al. Combining Feature Extraction Methods and Principal Component Analysis for Recognition of Vietnamese Off-Line Handwritten Uppercase Accented Characters
JPH08263591A (en) Device and method for character recognition
JPH09269970A (en) Method for recognizing character and its device
CN112966688A (en) Font detection method
JPS61163477A (en) Character recognition device
CN113887441A (en) Table data processing method, device, equipment and storage medium
JPS61206090A (en) Character reading device
CN111291726A (en) Medical bill sorting method, device, equipment and medium
JPH01209586A (en) Character recognizing system for sentence mixed with double size/half size characters
JP2004133841A (en) Pattern recognition device and pattern recognition method, program, and storage medium