Nothing Special   »   [go: up one dir, main page]

JPH09305713A - Pattern recognition system - Google Patents

Pattern recognition system

Info

Publication number
JPH09305713A
JPH09305713A JP8117324A JP11732496A JPH09305713A JP H09305713 A JPH09305713 A JP H09305713A JP 8117324 A JP8117324 A JP 8117324A JP 11732496 A JP11732496 A JP 11732496A JP H09305713 A JPH09305713 A JP H09305713A
Authority
JP
Japan
Prior art keywords
characters
recognition
dictionary
information
handwritten
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
JP8117324A
Other languages
Japanese (ja)
Inventor
Koichi Sato
浩一 佐藤
Masahito Fukushima
雅仁 福島
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.)
Hitachi Ltd
Original Assignee
Hitachi Ltd
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 Hitachi Ltd filed Critical Hitachi Ltd
Priority to JP8117324A priority Critical patent/JPH09305713A/en
Publication of JPH09305713A publication Critical patent/JPH09305713A/en
Pending legal-status Critical Current

Links

Landscapes

  • Character Discrimination (AREA)

Abstract

PROBLEM TO BE SOLVED: To recognize characters with high efficiency and at low cost by canceling partly or totally a handwrite information recognition routine against the written special characters and the writing way of a person who inputted the characters. SOLUTION: The characters inputted via a handwritten character input means 11 are expanded into various information including the vector, the writing order, the number of strokes, the character shapes, etc., by an input switching means 12. Based on these information, a handwritten character recognition means 13 recognizes characters via a recognition dictionary 15. A right-handed person writes the characters from the left to the right but a left-handed person is liable to write the characters from the right to the left. When a recognition means 18 inputs and designates the left-hander information, the vector information on the dictionary 15 is masked by a dictionary mask designation means 19, i.e., deleted out of the dictionary 15. Then the means 13 refers to the dictionary 15 and recognizes the handwritten characters. Under such conditions, no dictionary exists for the vector information and the recognition rate is improved even for the characters which are written by the left-handed person.

Description

【発明の詳細な説明】Detailed Description of the Invention

【0001】[0001]

【発明の属する技術分野】本発明は文字認識方式に関す
る。
TECHNICAL FIELD The present invention relates to a character recognition system.

【0002】[0002]

【従来の技術】近年、マルチメディア機器の発展によ
り、多種多様な機能を備えたパソコン,OA機器,マル
チメディア商品が増えてきており、パソコンの機能を搭
載し手軽に持ち運びが可能な携帯情報端末が普及してき
た。その携帯情報端末の入力手段として手書き文字入力
方式が主流となり、手書き文字認識率の向上が要求され
てきた。
2. Description of the Related Art In recent years, due to the development of multimedia equipment, the number of personal computers, office automation equipment, and multimedia products having a wide variety of functions has increased, and portable information terminals equipped with the functions of a personal computer and which can be carried easily. Has become popular. A handwritten character input method has become the mainstream as an input means of the portable information terminal, and it has been required to improve the handwritten character recognition rate.

【0003】従来の手書き文字認識方式は、通常の手書
き文字を認識する文字認識方式と、手書き文字から特定
のストローク群の有無を検出し文字認識する方式等、数
種類の手書き情報認識ルーチンを備え、手書き入力文字
を、この数種類の手書き情報認識ルーチンにより判定
し、手書き文字を認識していた。
The conventional handwritten character recognition system is provided with several types of handwritten information recognition routines such as a character recognition system for recognizing ordinary handwritten characters and a system for recognizing the presence or absence of a specific stroke group from the handwritten characters. The handwritten character was recognized by judging the handwritten input character by several kinds of handwriting information recognition routines.

【0004】図3に従来の文字認識方式を有するシステ
ム機能ブロックを示す。
FIG. 3 shows a system function block having a conventional character recognition system.

【0005】図3で、文字登録モードでは、登録手段3
7の制御により、手書き文字認識手段31から入力切り
替え手段32を通して手書き文字が出力され、入力要素
指定手段36から入力要素(文字または文字列)が出力
され、その手書き文字情報と入力要素情報の両方が手書
き文字を認識するときの辞書として認識辞書35の登録
領域に登録される。
In FIG. 3, in the character registration mode, the registration means 3
Under the control of 7, the handwritten character recognition unit 31 outputs the handwritten character through the input switching unit 32, and the input element designating unit 36 outputs the input element (character or character string), and both the handwritten character information and the input element information are output. Is registered in the registration area of the recognition dictionary 35 as a dictionary for recognizing handwritten characters.

【0006】手書き文字入力手段31から入力された文
字は、入力切り替え手段32により手書き文字認識手段
33に送られ、固定領域と登録領域からなる認識辞書3
5を参照して認識が行われ、その認識結果がコード変換
手段34を通じて出力される。
The characters input from the handwritten character input means 31 are sent to the handwritten character recognition means 33 by the input switching means 32, and the recognition dictionary 3 consisting of a fixed area and a registered area.
5, the recognition is performed, and the recognition result is output through the code converting means 34.

【0007】[0007]

【発明が解決しようとする課題】従来の手書き文字認識
方式は、数種類の手書き情報認識ルーチンを備え、手書
き入力文字を、この数種類の手書き情報認識ルーチンに
より判定し、手書き文字を認識していた。
The conventional handwritten character recognition system has several kinds of handwritten information recognition routines, and judges handwritten input characters by these several kinds of handwritten information recognition routines to recognize handwritten characters.

【0008】しかし、高度な手書き文字認識を行うため
に、この手書き情報認識ルーチンが多くなり、手書き文
字を認識するための手書き情報認識ルーチンが逆に、文
字認識の妨げになることがでてきた。
However, in order to perform high-level handwritten character recognition, the number of handwritten information recognition routines increases, and the handwritten information recognition routine for recognizing handwritten characters, on the contrary, can hinder character recognition. .

【0009】本発明の目的はこの多くの手書き情報認識
ルーチンによる誤認識を防ぐ手段を提供することにあ
る。
An object of the present invention is to provide a means for preventing erroneous recognition due to this many handwritten information recognition routines.

【0010】[0010]

【課題を解決するための手段】上記目的を達成するため
に、手書き情報認識ルーチンの一部または全部を解除可
能とすることを特徴とする。
In order to achieve the above object, a part or all of a handwriting information recognition routine can be released.

【0011】これにより手書き文字認識を行う際、この
手書き情報認識ルーチンの一部または全部を解除するこ
とによって手書き文字の認識率を高めることができる。
Thus, when performing handwritten character recognition, the recognition rate of handwritten characters can be increased by canceling a part or all of the handwriting information recognition routine.

【0012】上記文字認識方式は、使用者が手書き情報
認識ルーチンの一部または全部を解除することにより、
左利き右利き,使用者のくせ等による誤認式を防ぎ、手
書き文字の認識率を高めることができる。
In the above character recognition method, when the user cancels a part or all of the handwriting information recognition routine,
The recognition rate of handwritten characters can be improved by preventing misrecognition expressions due to left-handed right-handed users and habits of users.

【0013】[0013]

【発明の実施の形態】以下、本発明の一実施例について
図を用い説明する。
BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of the present invention will be described below with reference to the drawings.

【0014】図2は、本発明が適用される情報処理装置
の構成を示したものである。
FIG. 2 shows the configuration of an information processing apparatus to which the present invention is applied.

【0015】情報処理装置は、情報処理装置本体20で
あり、プログラムを実行するCPU21,実行するプログラム
や辞書等を格納するROM22,登録辞書として任意設
定値を格納するSRAM23,プログラムを実行するときにワ
ークエリアとして用いられるメインRAM24,入出力
用のI/Oポート25,タブレットを制御するタブレッ
ト制御部26,ディスプレイを制御するディスプレイ制
御部27,ディスプレイへの表示データを格納するVRAM
28等から構成される。一体型液晶パネル30はタブレッ
ト制御部26に接続されるタブレット21,ディスプレ
イ制御部27に接続される液晶ディスプレイ22などで
構成されている。
The information processing apparatus is the information processing apparatus main body 20, and includes a CPU 21 for executing a program, a ROM 22 for storing a program and a dictionary to be executed, an SRAM 23 for storing an arbitrary set value as a registered dictionary, and a program for executing the program. Main RAM 24 used as a work area, I / O port 25 for input / output, tablet control unit 26 for controlling the tablet, display control unit 27 for controlling the display, VRAM for storing display data to the display
Composed of 28 mag. The integrated liquid crystal panel 30 includes a tablet 21 connected to the tablet controller 26, a liquid crystal display 22 connected to the display controller 27, and the like.

【0016】この実施例で、図1に示す手書き文字入力
手段11はタブレット31により実現される。また、認
識辞書25の登録領域はSRAMにより実現され、固定
領域はROM22により実現される。他の各手段はプロ
グラム等で実現される。
In this embodiment, the handwritten character input means 11 shown in FIG. 1 is realized by the tablet 31. The registration area of the recognition dictionary 25 is realized by SRAM, and the fixed area is realized by ROM 22. Each other means is realized by a program or the like.

【0017】手書き文字入力手段11より入力した文字
は入力切り替え手段12によりベクトル情報,筆順,画
数,形状等の諸情報に展開する。これらの情報を手書き
文字認識手段13により、認識辞書15を参照し文字認
識を行う。このとき、認識辞書15にはベクトル情報,
筆順,画数,形状等の諸情報の辞書である認識辞書
(a)から認識辞書(z)、および個人がくせ字等の情
報を登録する登録辞書(a)から登録辞書(z)が格納
されており、これらの認識辞書15により諸情報を判断
し、文字認識を行うが、この際、認識辞書(a)から認
識辞書(z)、および登録辞書(a)から登録辞書
(z)の全てで文字認識を行うと認識率が下がる場合が
ある。例えば右利きの人が文字を書く場合、左から右へ
筆を運ぶのが普通であるが、左利きの人が文字を書く場
合は、筆を右から左へ運んでしまうため、認識辞書にベ
クトル情報をもって認識すると誤認識する可能性が高く
なる。そこで、認識指定手段18により、左利きの情報
を入力指定すると、辞書マスク指定手段19によって認
識辞書のベクトル情報をマスクされ、認識辞書15から
削除された形になり、手書き文字認識手段13によって
認識辞書15を参照,認識した場合、ベクトル情報の辞
書がなく左利きの人が書いた文字でも認識率が向上す
る。
The character input from the handwritten character input means 11 is expanded by the input switching means 12 into various information such as vector information, stroke order, number of strokes, and shape. The handwriting character recognizing means 13 refers to the recognition dictionary 15 for these pieces of information to perform character recognition. At this time, the recognition dictionary 15 stores vector information,
A recognition dictionary (z) is stored from a recognition dictionary (a) that is a dictionary of various information such as stroke order, stroke count, and shape, and a registration dictionary (z) is stored from a registration dictionary (a) in which an individual registers information such as cursive characters. The recognition dictionary 15 determines various information and performs character recognition. At this time, all of the recognition dictionary (a) to the recognition dictionary (z) and the registration dictionary (a) to the registration dictionary (z) are included. The recognition rate may decrease if you perform character recognition in. For example, when a right-handed person writes letters, it is common to carry the brush from left to right, but when a left-handed person writes letters, the brush is carried from right to left. If the information is recognized, the possibility of misrecognition increases. Therefore, when the recognition specifying means 18 inputs and specifies left-handed information, the dictionary mask specifying means 19 masks the vector information of the recognition dictionary and deletes it from the recognition dictionary 15, and the handwriting character recognition means 13 recognizes the recognition dictionary. When referring to 15 and recognizing, the recognition rate is improved even for characters written by a left-handed person without a dictionary of vector information.

【0018】[0018]

【発明の効果】本発明によれば、文字の入力者の特殊な
文字,書き方に対し、手書き情報認識ルーチンの一部ま
たは全部を解除することにより、安価でかつ効率良く文
字の認識を行うことができる。
According to the present invention, for a special character or writing style of a character inputting person, a part or all of the handwriting information recognition routine is canceled so that the character can be recognized inexpensively and efficiently. You can

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

【図1】本発明の一実施例である原理ブロック図、FIG. 1 is a block diagram showing the principle of an embodiment of the present invention,

【図2】本発明の一実施例であるハードウェアのブロッ
ク図。
FIG. 2 is a block diagram of hardware that is an embodiment of the present invention.

【図3】従来の文字認識方式のブロック図。FIG. 3 is a block diagram of a conventional character recognition method.

【符号の説明】[Explanation of symbols]

11…手書き文字入力手段、12…入力切り替え手段、
13…手書き文字認識手段、14…コード変換手段、1
5…認識辞書、16…入力要素指定手段、17…登録手
段、18…認識要素指定手段、19…辞書マスク指定手
段。
11 ... handwritten character input means, 12 ... input switching means,
13 ... Handwritten character recognition means, 14 ... Code conversion means, 1
5 ... Recognition dictionary, 16 ... Input element designation means, 17 ... Registration means, 18 ... Recognition element designation means, 19 ... Dictionary mask designation means.

Claims (1)

【特許請求の範囲】[Claims] 【請求項1】手書き文字入力手段と手書き文字認識手段
とを有するコンピュータシステムにおいて、文字認識す
るための種々の認識条件の一部または全部を解除可能と
することを特徴とする文字認識方式。
1. A character recognition system, wherein in a computer system having a handwritten character input means and a handwritten character recognition means, a part or all of various recognition conditions for character recognition can be canceled.
JP8117324A 1996-05-13 1996-05-13 Pattern recognition system Pending JPH09305713A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP8117324A JPH09305713A (en) 1996-05-13 1996-05-13 Pattern recognition system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP8117324A JPH09305713A (en) 1996-05-13 1996-05-13 Pattern recognition system

Publications (1)

Publication Number Publication Date
JPH09305713A true JPH09305713A (en) 1997-11-28

Family

ID=14708931

Family Applications (1)

Application Number Title Priority Date Filing Date
JP8117324A Pending JPH09305713A (en) 1996-05-13 1996-05-13 Pattern recognition system

Country Status (1)

Country Link
JP (1) JPH09305713A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2009506464A (en) * 2005-08-29 2009-02-12 マイクロソフト コーポレーション Use with handwriting input style

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2009506464A (en) * 2005-08-29 2009-02-12 マイクロソフト コーポレーション Use with handwriting input style

Similar Documents

Publication Publication Date Title
US5544262A (en) Method and apparatus for processing graphically input equations
CA2089784A1 (en) Apparatus and method for disambiguating an input stream generated by a stylus-based user interface
US5828783A (en) Apparatus and method for input-processing hand-written data
EP0459746A2 (en) Pattern recognition method, and apparatus therefor
CA2019132A1 (en) Object recognition system
JP2003196593A (en) Character recognizer, method and program for recognizing character
JPH09305713A (en) Pattern recognition system
Huerst et al. Interactive error repair for an online handwriting interface
JPS621086A (en) Character input device
JPH0442316A (en) Electronic computer
JPS59188787A (en) Input device of on-line manuscript character
Dutta et al. Lightweight user-adaptive handwriting recognizer for resource constrained handheld devices
JPH07320001A (en) Character recognition device and method for referring to recognition dictionary
JP2905334B2 (en) Online handwritten character recognition dictionary creation method and online handwritten character recognition dictionary creation device
JPH0634259B2 (en) Character recognition device
JPH02157986A (en) Recognizing character type designating system
JP2682395B2 (en) Character input device
JPH01145789A (en) Hand writing input device with user dictionary registering function
JPH09212080A (en) Portable information terminal device
KR100285429B1 (en) Method of after-processing character
JPS6084683A (en) Character recognizing system
JPS62190581A (en) Dictionary generation system
JPS61193282A (en) Hand-written character recognizing device
JPS6413682A (en) Handwriting information processor
JPH0573724A (en) Hand-written character recognition device