JPS6361388A - Histogram preparing method for character recognizing device - Google Patents

Histogram preparing method for character recognizing device

Info

Publication number
JPS6361388A
JPS6361388A JP61206638A JP20663886A JPS6361388A JP S6361388 A JPS6361388 A JP S6361388A JP 61206638 A JP61206638 A JP 61206638A JP 20663886 A JP20663886 A JP 20663886A JP S6361388 A JPS6361388 A JP S6361388A
Authority
JP
Japan
Prior art keywords
histogram
blocks
block
character
overlapping
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
JP61206638A
Other languages
Japanese (ja)
Inventor
Masahiro Nakamura
昌弘 中村
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.)
Ricoh Co Ltd
Original Assignee
Ricoh Co 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 Ricoh Co Ltd filed Critical Ricoh Co Ltd
Priority to JP61206638A priority Critical patent/JPS6361388A/en
Publication of JPS6361388A publication Critical patent/JPS6361388A/en
Pending legal-status Critical Current

Links

Landscapes

  • Character Discrimination (AREA)

Abstract

PURPOSE:To especially improve the recognition rate of a 7-segment character by dividing a character into plural blocks, overlapping a part of plural divided blocks to obtain a histogram. CONSTITUTION:A direction code is given to the contour part of a 7-segment character read from a scanner and the character is segmented into plural blocks 1-16. Next, the segmented blocks 1-16 are made into one block by overlapping a 2-block 1, 2, 2 and 3, 3 and 4...15 and 16 in an X direction, the code is counted by each direction to respective 12-overlapped blocks and a histogram is obtained. Next, one block is made by overlapping 2-block 1 and 5, 5 and 9...12 and 16 in the Y direction and the histogram is obtained. To the feature quantity of 24-block X 8 = 192 dimensions, the comparison operation is executed with a dictionary prepared for beforehand and a candidate character is obtained.

Description

【発明の詳細な説明】 技術分野 本発明は、音声認識装置におけるヒストグラム作成方法
に関する。
DETAILED DESCRIPTION OF THE INVENTION TECHNICAL FIELD The present invention relates to a histogram creation method in a speech recognition device.

史釆揉監 領域を分割し、その分割したブロック毎に方向別のコー
ドのヒストグラムをとり、このヒストグラムを特徴とし
て辞書とのマツチングを行なう7セグメント文字用のO
CRにおいて、第7図(b)に一般例として示したヒス
トグラムが文字が変形するとコードが入るブロックが変
わるためヒストグラ11の値が第7図(a)に示すよう
に変化し、距離が大きくなる誤認したりジェクトになっ
たりする。
The history control area is divided, a histogram of codes for each direction is taken for each divided block, and this histogram is used as a feature for matching with a dictionary.
In CR, the histogram shown as a general example in Figure 7(b) changes when the character changes, the block where the code is placed changes, so the value of histogram 11 changes as shown in Figure 7(a), and the distance increases. misidentify or become a subject.

目     的 本発明は、上述のごとき実情に鑑みてなされたもので、
特に、7セグメント文字の認識率を向上させることを目
的としてなされたものである。
Purpose The present invention was made in view of the above-mentioned circumstances.
In particular, this method was designed to improve the recognition rate of 7-segment characters.

構成 本発明は、上記目的を達成するために、スキャナーから
読み取った7セグメント文字の輪郭部に方向コードをっ
け、文字を複数のブロックに分割し、その分割した各々
のブロック毎に方向別のヒストグラムをとり、このヒス
トグラムを特徴量としてあらかじめ用意した辞書とマツ
チングを行なって文字を決定する文字認識装置において
、複数に分割したブロックの一部を重ねてヒストグラム
をとることを特徴としたものである。以下、本発明の実
施例に基づいて説明する。
Structure In order to achieve the above object, the present invention attaches a direction code to the outline of a 7-segment character read by a scanner, divides the character into a plurality of blocks, and assigns a direction code to each divided block. A character recognition device that determines characters by taking a histogram and matching this histogram with a dictionary prepared in advance as a feature quantity, which is characterized by overlapping parts of blocks divided into multiple blocks and taking a histogram. . Hereinafter, the present invention will be explained based on examples.

第1図は1本発明の詳細な説明するためのフローチャー
トすなわち領域分割を行なうフローチャートで、まず入
カバターンの輪郭部に付いた方向別のコードをカウント
し、その総数を求める。
FIG. 1 is a flowchart for explaining the present invention in detail, that is, a flowchart for performing area division.First, the codes attached to the outline of the input pattern in each direction are counted, and the total number thereof is determined.

次にX方向、Y方向への分割座標を求める。例えば、領
域をnXmに分割するとしてコードの総数  2S をSとした時、分割点はそれぞれ□、□、・・・n そこで画像をX方向、Y方向にスキャンし、それぞれコ
ードの数が各分割点になる座標を求める。
Next, the division coordinates in the X direction and Y direction are determined. For example, if the area is divided into nXm and the total number of codes 2S is S, the division points are □, □,...n. Then, the image is scanned in the X direction and the Y direction, and the number of codes is Find the coordinates of a point.

これが各ブロックの分割座標である。This is the division coordinate of each block.

第2図に、4×4に画像を分割した「8」の例を示す。FIG. 2 shows an example of "8" in which the image is divided into 4×4.

第3図は、ヒストグラム作成のフローチャート、第4図
は、領域を4×4のブロックに分割した場合の例を示す
図で、以下、この第4図を参照しながら、本発明による
ヒストグラム作成方法について説明する。まず、第4図
に示したブロックを(1,2)、(2,3)、(3,4
)、・・・(15,16)とX方向に重ねて2ブロツク
を1ブロツクとする。
FIG. 3 is a flowchart for creating a histogram, and FIG. 4 is a diagram showing an example of dividing an area into 4×4 blocks. I will explain about it. First, the blocks shown in Fig. 4 are (1, 2), (2, 3), (3, 4
), . . . (15, 16) are overlapped in the X direction to form one block.

この重ねた各12ブロツクに対して各方向別にコードを
カウントしヒストグラムをとる。
For each of the 12 stacked blocks, the codes are counted in each direction and a histogram is taken.

次に(1,5)、(5,9)・・・(12,16)とY
方向に重ねてヒストグラムをとる。この24ブロックX
8=192次元の特徴量に対して予め用意した辞書と比
較演算を行ない候補文字を求める。
Next, (1,5), (5,9)...(12,16) and Y
Take a histogram by overlapping in the direction. These 24 blocks
Candidate characters are obtained by performing a comparison operation on the 8=192-dimensional feature amounts with a dictionary prepared in advance.

このようにブロックを重ねてヒストグラムを作成するこ
とにより、第5図(a)、(b)において丸印A及びB
にて示すように、方向コードのつくブロックがずれてい
てもこれを吸収できるため比較演算による距離の増大を
防止でき、リジェクトや誤認を防ぐことができる。
By creating a histogram by overlapping blocks in this way, the circles A and B in Figures 5(a) and (b)
As shown in , even if the block with the direction code is misaligned, this can be absorbed, so it is possible to prevent the distance from increasing due to comparison calculations, and it is possible to prevent rejections and misidentifications.

第6図は1本発明の他の実施例によるヒストグラム作成
のフローチャートで、この実施例においては、重ね合わ
せたブロックによるヒストグラムとは別にもともとのブ
ロックのヒストグラムをとり、24X8+16X8=3
24次元の特徴量に対して辞書との演算を行なう。
FIG. 6 is a flowchart for creating a histogram according to another embodiment of the present invention. In this embodiment, a histogram of the original block is taken separately from the histogram of the superimposed blocks, and 24X8+16X8=3
Calculations with the dictionary are performed on 24-dimensional feature quantities.

効   果 以上の説明から明らかなように、本発明によると、X方
向、Y方向にブロックを重ね合わせてヒストグラムを作
成することにより、画像の上下左右の変形及び雑音によ
る方向コードの発生するブロックの変化を吸収すること
ができるので、リジェクト、誤認を防止することができ
る。
Effects As is clear from the above explanation, according to the present invention, by creating a histogram by overlapping blocks in the X and Y directions, it is possible to eliminate vertical and horizontal deformation of the image and blocks where directional codes occur due to noise. Since changes can be absorbed, rejections and misunderstandings can be prevented.

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

第1図は、領域分割の例を示すフローチャート。 第2図は、「8」を4×4に画像分割した例を示す図、
第3図は、ヒストグラム作成の例を示すフローチャート
、第4図は、領域を4×4のブロックに分割した場合の
例を示す図、第5図は、本発明によって作成したヒスト
グラムの例を示す図、第6図は、ヒストグラム作成方法
の他の例を示すフローチャート図、第7図は、従来のヒ
ストグラムの例を示す図である。 第1図 第3rX!  第6図 第2図 第7図 (a)                      
             (b)Oシ −ど2222
−、−−。
FIG. 1 is a flowchart showing an example of region division. FIG. 2 is a diagram showing an example of dividing the image of "8" into 4×4,
FIG. 3 is a flowchart showing an example of creating a histogram, FIG. 4 is a diagram showing an example of dividing an area into 4×4 blocks, and FIG. 5 is an example of a histogram created according to the present invention. FIG. 6 is a flowchart showing another example of a histogram creation method, and FIG. 7 is a diagram showing an example of a conventional histogram. Figure 1 Figure 3rX! Figure 6 Figure 2 Figure 7 (a)
(b) O-seed 2222
−、−−.

Claims (2)

【特許請求の範囲】[Claims] (1)、スキャナーから読み取った7セグメント文字の
輪郭部に方向コードをつけ、文字を複数のブロックに分
割し、その分割した各々のブロック毎に方向別のヒスト
グラムをとり、このヒストグラムを特徴量としてあらか
じめ用意した辞書とマッチングを行なって文字を決定す
る文字認識装置において、複数に分割したブロックの一
部を重ねてヒストグラムをとることを特徴とする文字認
識装置におけるヒストグラム作成方法。
(1) Add a direction code to the outline of the 7-segment character read by the scanner, divide the character into multiple blocks, take a histogram for each direction for each divided block, and use this histogram as a feature quantity. A method for creating a histogram in a character recognition device, which determines a character by matching with a dictionary prepared in advance, and is characterized in that a histogram is obtained by overlapping parts of blocks divided into a plurality of blocks.
(2)、上記重ねたブロックのヒストグラムと重ねる前
のブロックのヒストグラムを組み合わせることを特徴と
する特許請求の範囲第(1)項に記載の文字認識装置に
おけるヒストグラム作成方法。
(2) A histogram creation method in a character recognition device according to claim (1), characterized in that the histogram of the overlapping block and the histogram of the block before overlapping are combined.
JP61206638A 1986-09-02 1986-09-02 Histogram preparing method for character recognizing device Pending JPS6361388A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP61206638A JPS6361388A (en) 1986-09-02 1986-09-02 Histogram preparing method for character recognizing device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP61206638A JPS6361388A (en) 1986-09-02 1986-09-02 Histogram preparing method for character recognizing device

Publications (1)

Publication Number Publication Date
JPS6361388A true JPS6361388A (en) 1988-03-17

Family

ID=16526672

Family Applications (1)

Application Number Title Priority Date Filing Date
JP61206638A Pending JPS6361388A (en) 1986-09-02 1986-09-02 Histogram preparing method for character recognizing device

Country Status (1)

Country Link
JP (1) JPS6361388A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0355748A2 (en) * 1988-08-20 1990-02-28 Fujitsu Limited A pattern recognition apparatus and method for doing the same

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0355748A2 (en) * 1988-08-20 1990-02-28 Fujitsu Limited A pattern recognition apparatus and method for doing the same

Similar Documents

Publication Publication Date Title
US4607384A (en) Fingerprint classification arrangement
JPS61188681A (en) Character recognition device
JPH06162198A (en) Pattern recognizing device
JPS6361388A (en) Histogram preparing method for character recognizing device
Hobby Polygonal approximations that minimize the number of inflections
JP2853169B2 (en) Pattern recognition device
JPS62192886A (en) Feature quantity generating method in character recognizing device
JPS613287A (en) Graphic form input system
JP2789622B2 (en) Character / graphic area determination device
JP2941322B2 (en) Drawing processing equipment
JPS636676A (en) Stroke instructing system
JP2988697B2 (en) Figure recognition method
JPS6334682A (en) Character recognition system
JPS6316791B2 (en)
JP2643294B2 (en) Dictionary creation method
JPH02227786A (en) Intersection extracting method in msm encoding system
JPS6295687A (en) Character recognizing system
JPH0632080B2 (en) Character recognition method
JPS6375989A (en) Picture recognizing system
JPS6174079A (en) Pattern recognizing device
JPH0215388A (en) Character recognizing device
JPH0695644A (en) System for recognizing straight line of character pattern
JPS6140685A (en) Character recognizing device
JPS5822777B2 (en) Optical character reading method
JPS5960582A (en) Optical character reader