JPH0223906B2 - - Google Patents
Info
- Publication number
- JPH0223906B2 JPH0223906B2 JP56070761A JP7076181A JPH0223906B2 JP H0223906 B2 JPH0223906 B2 JP H0223906B2 JP 56070761 A JP56070761 A JP 56070761A JP 7076181 A JP7076181 A JP 7076181A JP H0223906 B2 JPH0223906 B2 JP H0223906B2
- Authority
- JP
- Japan
- Prior art keywords
- dictionary
- reading
- characters
- character
- font
- 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 - Lifetime
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/24—Character recognition characterised by the processing or recognition method
- G06V30/242—Division of the character sequences into groups prior to recognition; Selection of dictionaries
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Character Discrimination (AREA)
Description
【発明の詳細な説明】
本発明はマルチフオントに適合した光学的文学
読取装置（以下OCRと略称する）に関する。DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a multi-font compatible optical literature reader (hereinafter abbreviated as OCR).
近年、活字読取OCRで多文字種読取のものが
増えているが、字種が増えるに従つて類似文字も
増加するため読取率が低下する傾向が強くなつて
来た。又ミツクスマルチ（同時混在読取）用の辞
書作成の面でも長時間大容量の計算機を使用しな
ければならず技術的にも問題があつた。 In recent years, the number of print reading OCRs that can read multiple character types has increased, but as the number of character types increases, the number of similar characters also increases, so there is a strong tendency for the reading rate to decline. In addition, creating a dictionary for mix-multi (simultaneous mixed reading) required the use of a large-capacity computer for a long time, which caused technical problems.
本発明は以上の問題点を解消できるOCRを提
供することを目的とするものである。 An object of the present invention is to provide an OCR that can solve the above problems.
本発明はフオント毎に独立して構成されるマル
チ辞書を使用し最初の数文字は全辞書を用いて読
取を行い最も相性の良い（平均的類似度の高い）
辞書を自動選択し、選択された辞書を用いて以降
の読取を行うようにしたものである。この場合、
見かけ上はミツクスマルチを読んでいるように見
えるが内容は大いに異なるものである。 The present invention uses a multi-dictionary that is configured independently for each font, and reads the first few characters using all dictionaries to find the most compatible (high average similarity)
A dictionary is automatically selected and subsequent reading is performed using the selected dictionary. in this case,
At first glance, it looks like you are reading a mix multi, but the content is very different.
以下に実施例により本発明の詳細を説明する。
図は本発明の一実施例の機能を示すブロツク図で
ある。同図において、１は字種毎に独立に構成さ
れた辞書をもつセレクテイブマルチフオント辞
書、２は文字の認識部であり、帳票の最初の一定
数の文字については全辞書を引いて読取を行い、
各文字について各辞書における最高の類似度値を
得点として独立に加算し、各累積点を得点記憶レ
ジスタ３に記憶させて行き、一定数の文字につい
てこの作業を繰返した後、比較器４を用いて最高
得点を検出し、該帳票で使用する辞書を自動選択
する機能をもつものである。 The details of the present invention will be explained below with reference to Examples.
The figure is a block diagram showing the functions of one embodiment of the present invention. In the figure, 1 is a selective multi-font dictionary that has dictionaries configured independently for each type of character, and 2 is a character recognition unit, which reads the first certain number of characters in a form by referring to the entire dictionary. and
For each character, the highest similarity value in each dictionary is independently added as a score, and each cumulative score is stored in the score storage register 3. After repeating this process for a certain number of characters, the comparator 4 is used to It has a function to detect the highest score and automatically select the dictionary to be used in the form.
前記した最初の一定数の文字の読取については
辞書選択の段階であり、この時点では答を出かず
に各辞書が得た最高の類似度値を辞書毎に加算し
て行き、一定文字数の処理が終つたところで各辞
書の類似度値の累積点を比較し、最高の累積点を
示した辞書を該帳票で使用することにし、第１文
字目から識別を再開する。 The reading of the first certain number of characters mentioned above is at the dictionary selection stage, and at this point, the highest similarity value obtained by each dictionary is added up for each dictionary, and the processing of a certain number of characters is performed. When this is completed, the cumulative points of the similarity values of each dictionary are compared, and the dictionary with the highest cumulative score is decided to be used for the form, and identification is restarted from the first character.
辞書選択後の読取に於て、リジエクトが多発
し、一定数値を超えた場合は、辞書選択の為のサ
ンプルを増やしてもう一度辞書の選択を行う。 During reading after dictionary selection, if rejects occur frequently and exceed a certain value, samples for dictionary selection are increased and dictionary selection is performed again.
後戻りして読取が行えない構造のOCRでは辞
書選択段階に於ても各辞書毎に答を保存してお
き、辞書決定時に対応するものを正式の答とし採
用することによつて同様な結果を得ることができ
る。 In OCR, which has a structure that does not allow for backward reading, it is possible to obtain similar results by saving the answer for each dictionary even at the dictionary selection stage, and by adopting the corresponding answer as the official answer when selecting the dictionary. Obtainable.
本発明は以上のようになるものであつて、(i)最
初の数文字については全辞書を使用するがそれ以
降は単独の辞書で読取るので、高精度、高速化
（ミツクスマルチに比べて）が可能である。(ii)辞
書の開発が容易である。等の効果が得られる。 The present invention is as described above, and (i) the first few characters are read using the entire dictionary, but after that, a single dictionary is used to read them, resulting in high accuracy and high speed (compared to mix multi). It is possible. (ii) It is easy to develop a dictionary. Effects such as this can be obtained.
図は本発明の一実施例の機能を示すブロツク図
である。
１：マルチフオント辞書、２：認識部、３：得
点記憶レジスタ、４：比較器。
The figure is a block diagram showing the functions of one embodiment of the present invention. 1: Multi-font dictionary, 2: Recognition unit, 3: Score storage register, 4: Comparator.
Claims (1)
れらの辞書を用いて先づ最初の一定数の文字の読
取りを行い各文字毎に各辞書から得た最高類似度
値を辞書毎に加算する手段と、この加算結果を比
較して使用する辞書を選択する手段と、この選択
された辞書に基づいて文字を読取りその読取りリ
ジエクトが所定限度を超えたとき再度辞書の選択
を行わせる手段とを具備したことを特徴とするマ
ルチフオント文字読取装置。1 A group of dictionaries configured independently for each font, and using these dictionaries, first read a certain number of characters first, and then add the highest similarity value obtained from each dictionary for each character for each dictionary. means for comparing the addition results and selecting a dictionary to be used; and means for reading characters based on the selected dictionary and causing dictionary selection to be performed again when the reading reject exceeds a predetermined limit. A multi-font character reading device characterized by:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP56070761A JPS57187780A (en) | 1981-05-13 | 1981-05-13 | Multifont character reader |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP56070761A JPS57187780A (en) | 1981-05-13 | 1981-05-13 | Multifont character reader |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| JPS57187780A JPS57187780A (en) | 1982-11-18 |
| JPH0223906B2 true JPH0223906B2 (en) | 1990-05-25 |
Family
ID=13440808
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP56070761A Granted JPS57187780A (en) | 1981-05-13 | 1981-05-13 | Multifont character reader |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JPS57187780A (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0219986A (en) * | 1988-07-08 | 1990-01-23 | Fujitsu Ltd | Method for determining adaptive dictionary in character recognition and character recognizing device to execute above-mentioned method |
| JP3131287B2 (en) * | 1992-05-27 | 2001-01-31 | 株式会社日立製作所 | Pattern recognition device |
| JP2021018470A (en) * | 2019-07-17 | 2021-02-15 | 東芝テック株式会社 | Article specification device and program |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS5410266B2 (en) * | 1973-06-29 | 1979-05-02 |
-
1981
- 1981-05-13 JP JP56070761A patent/JPS57187780A/en active Granted
Also Published As
| Publication number | Publication date |
|---|---|
| JPS57187780A (en) | 1982-11-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US4030068A (en) | Optical character recognition system | |
| JPH0223906B2 (en) | ||
| JP2503208B2 (en) | Business card image processing method | |
| JPH024033B2 (en) | ||
| JPS6139175A (en) | Optical character reading device | |
| JPS6464085A (en) | Slip format registering device | |
| JPS63782A (en) | Pattern recognizing device | |
| JP3217442B2 (en) | Optical character reader | |
| JPS59158482A (en) | Character recognizing device | |
| JP2746345B2 (en) | Post-processing method for character recognition | |
| JPS61251984A (en) | Device for recognizing multi-font type character | |
| JP2969751B2 (en) | Character recognition processing method | |
| JPS6118080A (en) | Character recognizer | |
| JPS63195783A (en) | Character segmenting system | |
| JPH07121665A (en) | Compiling method and retrieving method for character recognition dictionary | |
| JPS6473482A (en) | Method for recognizing character | |
| JPS6336487A (en) | Character reading system | |
| JPS57121755A (en) | Aerial photography processing system | |
| JPS61136180A (en) | Character recognizer | |
| JPS55146577A (en) | Character recognition system | |
| JPS5960685A (en) | Optical character reader | |
| JPH0414193A (en) | Italic character recognition method | |
| JPS5999586A (en) | Optical character reader | |
| JPH01183796A (en) | Character recognizing device | |
| JPH0496190A (en) | Device and method for post-processor for optical hand-written kanji |