JPS62249294A - Individual dictionary registering method - Google Patents
Individual dictionary registering methodInfo
- Publication number
- JPS62249294A JPS62249294A JP61092228A JP9222886A JPS62249294A JP S62249294 A JPS62249294 A JP S62249294A JP 61092228 A JP61092228 A JP 61092228A JP 9222886 A JP9222886 A JP 9222886A JP S62249294 A JPS62249294 A JP S62249294A
- Authority
- JP
- Japan
- Prior art keywords
- character
- user
- dictionary
- feature amount
- registered
- 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
Links
- 238000000034 method Methods 0.000 title claims description 8
- 238000003909 pattern recognition Methods 0.000 claims 1
- 101100328887 Caenorhabditis elegans col-34 gene Proteins 0.000 abstract description 7
- 230000004044 response Effects 0.000 abstract description 4
- 230000003247 decreasing effect Effects 0.000 abstract 1
- 238000000605 extraction Methods 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 1
Landscapes
- Character Discrimination (AREA)
Abstract
Description
【発明の詳細な説明】
■技術分野
本発明は、文字認識、音声認識等の個人辞書登録方法に
関するものである。DETAILED DESCRIPTION OF THE INVENTION Technical Field The present invention relates to a personal dictionary registration method for character recognition, voice recognition, etc.
■従来技術
従来１文字認識、音声認識等の辞書（４ｍ準パターン）
は、予め供給されたものを使用したり、ユーザが使用す
る全ての文字や単語を書いたり、発声したりすることに
より登録した個人辞書を用いている。■Prior art Dictionary for conventional single character recognition, voice recognition, etc. (4m quasi-pattern)
uses a personal dictionary that is either pre-supplied or registered by the user by writing or saying all the letters and words he or she wants to use.
ところが、漢字文字認識や、大語い音声認識では、その
対象文字や単語の数が数千にも及ぶことから１個人辞書
を作成した方が認識率は向上するにもかかわらず、ユー
ザに多大な負担をかけるために１個人辞書の登録が非常
に困難であった。However, in kanji character recognition and large-word speech recognition, the number of target characters and words is in the thousands, so although creating a personal dictionary improves the recognition rate, it costs the user a lot. Registration of a personal dictionary was extremely difficult due to the heavy burden imposed on the user.
■ 目的
本発明は、文字認識、音声認識等において、簡便でユー
ザの負担を軽減するような個人辞書登録システム及びそ
の登録方法を提供することを目的とするものである。(1) Purpose It is an object of the present invention to provide a personal dictionary registration system and its registration method that are simple and reduce the burden on the user in character recognition, voice recognition, etc.
■構成
第１図は本発明の一実施例の構成を示すブロック図であ
る。(2) Configuration FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention.
これは１画像を入力するスキャナー１と、文字認識部３
．辞書をロードするための辞書ＲＡＭ４及び文字認識部
３をコントロールするＯＣＲＣシコントロールより或る
認識装置２と、辞書登録システムコントロール部６．標
準辞書７．ディスプレイ８．キーボード９及びディスク
１０から或る。This consists of a scanner 1 that inputs one image, and a character recognition unit 3.
．． A recognition device 2 from an OCRC controller that controls a dictionary RAM 4 for loading a dictionary and a character recognition section 3, and a dictionary registration system control section 6. Standard dictionary 7. Display 8. From the keyboard 9 and disk 10.
第２図（ａ）及び（ｂ）は第１図の動作を説明するため
のフローチャートである。FIGS. 2(a) and 2(b) are flowcharts for explaining the operation of FIG. 1.
（実施例１）
最初は、予め供給された辞書を辞書ＲＡＭ４上にロード
して使用する。(Embodiment 1) Initially, a dictionary supplied in advance is loaded onto the dictionary RAM 4 and used.
認識処理中にリジェクトした文字やユーザが修正した文
字が発生したとき、あるいは、ユーザが直接指定した場
合に、辞書登録システムコントロール部６は、ユーザに
確認し、登録モードとなる。When a rejected character or a character modified by the user occurs during the recognition process, or when the user directly specifies the character, the dictionary registration system control unit 6 confirms with the user and enters the registration mode.
そして、ユーザが記入した１文字当り複数個の文字パタ
ーンを文字認識部３へ送り、特徴抽出を行い、その特徴
量ＡＶ、と、もともとのその文字に対する特徴量Ｃｋｉ
との距離、あるいは、類似度等がある閾値Ｔｈ、以上な
らばユーザに警告を発し、ユーザの応答が登録に可なら
ばＡＶ、をその文字の新しい辞書として辞書ＲＡＭ４に
書込む。Then, a plurality of character patterns for each character written by the user are sent to the character recognition unit 3, and features are extracted, and the feature amount AV and the original feature amount Cki for that character are extracted.
If the distance or similarity is greater than a certain threshold Th, a warning is issued to the user, and if the user's response is acceptable for registration, AV is written into the dictionary RAM 4 as a new dictionary for that character.
また、認識処理の最後にユーザが確認を取り、現在の辞
書ＲＡＭ４をユーザの個人辞書としてディスク１０上に
保存し、以後、そのユーザが使用する際には、ディスク
上から辞書を辞書ＲＡＭ４ヘロードする。Also, at the end of the recognition process, the user confirms, and the current dictionary RAM 4 is saved on the disk 10 as the user's personal dictionary, and from then on, when the user uses it, the dictionary is loaded from the disk to the dictionary RAM 4. .
第２図（ａ）において、 Ｎ：特徴量の次元数 ｃｋｉ：文字にの特徴量 である。In Figure 2(a), N: Number of dimensions of features cki: Characteristic amount of characters.
（実施例２）
実施例１において、ユーザが記入した文字パターンから
得られた平均特徴量ＡＶ、と、その文字以外の辞書ＲＡ
Ｍ４上の特徴量Ｃ，，どの距離、あるいは、類似度を全
て求め、その最小値（ｄ２）が、或る閾値Ｔｈ、以下な
らばユーザに警告を発し、ユーザの応答が登録に可なら
ばＡ Ｖ Ｉをその新しい辞書として辞書ＲＡＭ４に書
込む、゛第２図（ｂ）において。(Example 2) In Example 1, the average feature amount AV obtained from the character pattern entered by the user and the dictionary RA other than the character
Find all distances or similarities on M4, and if the minimum value (d2) is less than a certain threshold Th, issue a warning to the user, and if the user's response is acceptable for registration, AV I is written into the dictionary RAM 4 as its new dictionary. ``In FIG. 2(b).
（実施例３）
平均特徴量ＡＶ、と辞書ＲＡＭ上のその文字に対する特
徴量Ｃｋｌとの距離（ｄｉ）、あるいは、類似度が或る
閾値Ｔｈ１より小さく、かつ、その文字以外の特徴量Ｃ
，ｌとの距離、あるいは、類似度の最小値（ｄ８）が、
或る閾値Ｔｈ、より大きいときのみ辞書に登録し、それ
以外はユーザに警告を発し、ユーザの応答が登録に可な
らばＡＶ、を辞書に登録する。(Example 3) The distance (di) between the average feature amount AV and the feature amount Ckl for that character on the dictionary RAM, or the feature amount C for which the degree of similarity is smaller than a certain threshold Th1 and other than that character
, l, or the minimum value of similarity (d8) is
The AV is registered in the dictionary only when it is larger than a certain threshold Th, otherwise a warning is issued to the user, and if the user's response is acceptable for registration, the AV is registered in the dictionary.
■効果 以上説明したように、本発明は、文字認識装置。■Effect As explained above, the present invention provides a character recognition device.
音声認識装置等において１個人辞書を作成するのに全て
の文字、単語をユーザが登録するのではなく、ユーザが
装置を使用しながら個人辞書を徐々に整備して行くこと
が可能となり、ユーザの負担を軽減できる。Instead of the user having to register all the characters and words to create one personal dictionary using a voice recognition device, etc., it becomes possible for the user to gradually improve the personal dictionary while using the device. It can reduce the burden.
また、辞書に対する学習機能を装置に付加することが出
来るという利点がある。Further, there is an advantage that a dictionary learning function can be added to the device.
第１図は本発明の一実施例の構成を示すブロック図、第
２図（ａ）及び（ｂ）は第１図の動作を説明するための
フローチャートである。
１・・・スキャナー、２・・・認識装置、３・・・文字
認識部、４・・・辞書ＲＡＭ。
５・・・ＯＣＲコントロール部、
６・・・辞書登録システムコントロール部、７・・・標
準辞書、８・・・ディスプレイ。
９ ・・・キーボード、１０・・・ディスク。
第１図
第２図FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention, and FIGS. 2(a) and 2(b) are flowcharts for explaining the operation of FIG. 1. 1...Scanner, 2...Recognition device, 3...Character recognition section, 4...Dictionary RAM. 5...OCR control unit, 6...Dictionary registration system control unit, 7...Standard dictionary, 8...Display. 9...Keyboard, 10...Disk. Figure 1 Figure 2
Claims (3)
置において、 リジェクト文字、ユーザが修正した文字、ユーザが指示
した文字に対して、ユーザが記入した複数の文字から求
めた平均特徴量と、予め登録されていたその文字に対す
る特徴量とを比較し、新たにその特徴量を登録するか否
かを判断することを特徴とする個人辞書登録方法。(1) In a pattern recognition device such as a character recognition device or a voice recognition device, an average feature amount calculated from a plurality of characters written by the user for rejected characters, characters corrected by the user, and characters specified by the user; A personal dictionary registration method characterized by comparing the feature amount for the character that has been registered in advance and determining whether or not to newly register the feature amount.
外の辞書の特徴量とを比較し、距離の最小値が閾値以下
かどうか、あるいは、類似度の最大値が或る閾値以上か
どうかで、新たにその特徴量を登録するか否かを判断す
ることを特徴とする特許請求の範囲第（１）項記載の個
人辞書登録方法。(2) Compare the average feature obtained from multiple characters with the dictionary features other than that character, and check whether the minimum distance is less than a threshold, or the maximum similarity is more than a certain threshold The personal dictionary registration method according to claim (1), wherein it is determined whether or not to newly register the feature amount based on whether or not the feature amount is newly registered.
対する特徴量とを比較し、かつ、平均特徴量と、その文
字以外の辞書の特徴量とを比較することにより、新たに
その特徴量を登録するか否かを判断することを特徴とす
る特許請求の範囲第（１）項記載の個人辞書登録方法。(3) By comparing the average feature amount obtained from multiple characters with the feature amount for that character, and by comparing the average feature amount with the feature amount of the dictionary other than that character, new features can be obtained. A personal dictionary registration method according to claim (1), characterized in that it is determined whether or not to register a quantity.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP61092228A JPS62249294A (en) | 1986-04-23 | 1986-04-23 | Individual dictionary registering method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP61092228A JPS62249294A (en) | 1986-04-23 | 1986-04-23 | Individual dictionary registering method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| JPS62249294A true JPS62249294A (en) | 1987-10-30 |
Family
ID=14048577
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP61092228A Pending JPS62249294A (en) | 1986-04-23 | 1986-04-23 | Individual dictionary registering method |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JPS62249294A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH03135598A (en) * | 1989-10-20 | 1991-06-10 | Matsushita Electric Ind Co Ltd | Character font output device |
| EP0667590A2 (en) * | 1994-02-10 | 1995-08-16 | Canon Kabushiki Kaisha | Method of registering a character pattern into a user dictionary and a character recognition apparatus having the user dictionary |
| JP2002074269A (en) * | 2000-08-28 | 2002-03-15 | Nippon Digital Kenkyusho:Kk | Method for recognizing character |
| JP2002182685A (en) * | 2000-12-12 | 2002-06-26 | Sony Corp | Recognizer and recognition system, learning system and learning method as well as recording medium |
-
1986
- 1986-04-23 JP JP61092228A patent/JPS62249294A/en active Pending
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH03135598A (en) * | 1989-10-20 | 1991-06-10 | Matsushita Electric Ind Co Ltd | Character font output device |
| EP0667590A2 (en) * | 1994-02-10 | 1995-08-16 | Canon Kabushiki Kaisha | Method of registering a character pattern into a user dictionary and a character recognition apparatus having the user dictionary |
| EP0667590A3 (en) * | 1994-02-10 | 1996-05-08 | Canon Kk | Method of registering a character pattern into a user dictionary and a character recognition apparatus having the user dictionary. |
| US5754686A (en) * | 1994-02-10 | 1998-05-19 | Canon Kabushiki Kaisha | Method of registering a character pattern into a user dictionary and a character recognition apparatus having the user dictionary |
| JP2002074269A (en) * | 2000-08-28 | 2002-03-15 | Nippon Digital Kenkyusho:Kk | Method for recognizing character |
| JP4580520B2 (en) * | 2000-08-28 | 2010-11-17 | 株式会社日本デジタル研究所 | Character recognition method and character recognition apparatus |
| JP2002182685A (en) * | 2000-12-12 | 2002-06-26 | Sony Corp | Recognizer and recognition system, learning system and learning method as well as recording medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JPS62249294A (en) | Individual dictionary registering method | |
| JPS58214973A (en) | Similar character discriminating system | |
| JPH01125683A (en) | Character recognizing device | |
| JPS5929246Y2 (en) | Online recognition processing device for handwritten characters | |
| JPS6224382A (en) | Method for recognizing handwritten character | |
| JP2938276B2 (en) | Character recognition device | |
| JP2840281B2 (en) | Character processing apparatus and method | |
| JPH06131492A (en) | License plate recognizing method | |
| JP2002259978A (en) | Pattern recognition device | |
| JPH01180083A (en) | Device for recognizing character of plural font | |
| JPH03126188A (en) | Character recognizing device | |
| JPH07160820A (en) | Character recognizing method | |
| JPS61100884A (en) | Character recognizing device | |
| JP4143148B2 (en) | Character recognition device | |
| JPH0313630B2 (en) | ||
| JPH0437467B2 (en) | ||
| JPH0344355B2 (en) | ||
| JPH09128484A (en) | Character recognizing method | |
| JPH034953B2 (en) | ||
| JPS61131184A (en) | Pattern recognizing device | |
| JPH0573027A (en) | Individual penmanship dictionary generation device and character output processor using individual penmanship dictionary | |
| JPH0676061A (en) | Dictionary preparing method for pattern recognition | |
| JP2001143020A (en) | Character recognition device, method and recording medium thereof | |
| JPH0229885A (en) | Character segmenting system | |
| JPH053633B2 (en) |