JPH07104940B2 - Figure recognition device - Google Patents

Figure recognition device

Info

Publication number
JPH07104940B2
JPH07104940B2 JP62238413A JP23841387A JPH07104940B2 JP H07104940 B2 JPH07104940 B2 JP H07104940B2 JP 62238413 A JP62238413 A JP 62238413A JP 23841387 A JP23841387 A JP 23841387A JP H07104940 B2 JPH07104940 B2 JP H07104940B2
Authority
JP
Japan
Prior art keywords
information
graphic
image
symbol
isolated
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 - Fee Related
Application number
JP62238413A
Other languages
Japanese (ja)
Other versions
JPS6481081A (en
Inventor
聡 田中
秀明 金近
剛 七野
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.)
Mitsubishi Electric Corp
Original Assignee
Mitsubishi Electric 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 Mitsubishi Electric Corp filed Critical Mitsubishi Electric Corp
Priority to JP62238413A priority Critical patent/JPH07104940B2/en
Publication of JPS6481081A publication Critical patent/JPS6481081A/en
Publication of JPH07104940B2 publication Critical patent/JPH07104940B2/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Landscapes

  • Image Analysis (AREA)

Description

【発明の詳細な説明】 〔産業上の利用分野〕 この発明は図面に書かれている情報を自動的に計算機等
へ入力するために,図面中の文字や孤立シンボルを検出
して認識する図形認識装置に関するものである。
DETAILED DESCRIPTION OF THE INVENTION [Industrial field of use] The present invention detects a character or an isolated symbol in a drawing and recognizes it in order to automatically input the information written in the drawing into a computer or the like. The present invention relates to a recognition device.

〔従来の技術〕[Conventional technology]

第10図は例えば電気学会雑誌昭和59年4月(104巻4
号)の図面読取技術の動向に示されている従来の図形認
識装置である。図において,(1)は図面を読取つて2
値の画像情報を作成する画像入力手段,(2)は画像入
力手段(1)によつて作成された画像情報を記憶する画
像記憶手段,(11)はこの画像記憶手段(2)によつて
記憶されている画像情報を入力し,画像情報中の塗りつ
ぶし領域等の情報を抽出し,線の太さが1になるまで線
を細める細線化処理を行ない,線の接続関係等の構造情
報をとり出す画像処理手段,(12)はこの画像処理手段
(11)によつて作成された線や領域の情報を記憶する線
情報記憶手段,(13)は線情報記憶手段(12)によつて
記憶されている線や領域の情報を入力して文字やシンボ
ルの候補となる線や領域情報をとり出す図形要素分離手
段,(14)は図形要素分離手段(13)によつてとり出さ
れた文字,シンボル候補の線,領域情報を記憶する図形
要素記憶手段,(15)は図形要素記憶手段(14)によつ
て記憶されている線,領域情報と画像記憶手段(2)に
よつて記憶されている画像情報を入力して,シンボルの
種類を認識するシンボル認識手段,(8)はシンボル認
識手段(15)によつて認識されたシンボルコードを記憶
するシンボル記憶手段,(9)は図形要素記憶手段(1
4)によつて記憶されている線,領域情報と画像記憶手
段(2)によつて記憶されている画像情報を入力して文
字を認識する文字認識手段,(10)は文字認識手段
(9)によつて認識された文字コードを記憶する文字記
憶手段である。
Figure 10 shows, for example, the Institute of Electrical Engineers, April 1984 (Vol. 104, No. 4).
No.), a conventional figure recognition device shown in the trend of drawing reading technology. In the figure, (1) is for reading the drawing
Image input means for creating value image information, (2) image storage means for storing the image information created by the image input means (1), and (11) by this image storage means (2) Input the stored image information, extract the information such as the filled area in the image information, perform the thinning process to thin the line until the thickness of the line becomes 1, and obtain the structural information such as the connection relation of the line. An image processing means for taking out, (12) a line information storing means for storing information on a line or an area created by the image processing means (11), and (13) a line information storing means (12) The graphic element separating means (14) for extracting the line and area information which is a candidate for the character or symbol by inputting the stored information about the line and area is extracted by the graphic element separating means (13). Characters, lines of symbol candidates, graphic element storage means for storing area information, (15) Symbol recognition means for recognizing the type of symbol by inputting the line and area information stored by the graphic element storage means (14) and the image information stored by the image storage means (2), ( 8) is a symbol storage means for storing the symbol code recognized by the symbol recognition means (15), and (9) is a graphic element storage means (1
4) is a character recognition means for recognizing a character by inputting line and area information stored by the image storage means (2) and image information stored by the image storage means (2), and (10) is a character recognition means (9). ) Is a character storage means for storing the character code recognized by the.

従来の図形認識装置は上記のように構成され,図面を画
像入力手段(1)により入力すると,太さをもつた2値
のデイジタル画像データとして,画像記憶手段(2)に
格納される。画像処理手段(11)は画像データより,孤
立点や穴点の雑音除去及び塗りつぶし領域の情報を抽出
後,線の太さが1になるまで細線化し,線の中心の座標
列と線間の接続関係をもつ線情報をとり出し,線情報記
憶手段(12)に格納する。図形要素分離手段(13)は,
とりだされた線情報に対し,例えば連結した線の大きさ
が一定以下の線の集まりや小ループや塗りつぶし,こま
かい折れ曲がり点のある線等々をシンボル,文字候補と
なる図形要素としてとり出し,図形要素記憶手段(14)
に格納する。シンボル認識手段(15)は分離されたシン
ボル候補となる図形要素の情報から画像記憶手段(2)
に格納されているもとの画像情報を切り出し,シンボル
か否かを判定し,シンボルと判定されたものに対してシ
ンボルコードをシンボル記憶手段(8)に格納する。文
字記憶手段(9)も分離された文字候補となる一定の大
きさの図形要素の情報から,画像記憶手段(2)に格納
されているもとの画像を切り出し,文字か否かを判定
し,文字と判定されたものは文字コードを文字記憶手段
(10)に格納する。
The conventional graphic recognition apparatus is configured as described above, and when a drawing is input by the image input means (1), it is stored in the image storage means (2) as binary digital image data having a thickness. The image processing means (11) removes noise of isolated points and hole points from the image data and extracts information of the filled area, and then thins the line until the thickness of the line becomes 1, and the coordinate sequence at the center of the line and the space between the lines. The line information having a connection relationship is extracted and stored in the line information storage means (12). The graphic element separating means (13) is
For the extracted line information, for example, a set of connected lines whose sizes are less than a certain size, small loops, fills, lines with fine bending points, etc. are extracted as symbols and graphic elements as character candidates. Element storage means (14)
To store. The symbol recognition means (15) uses the information of the separated graphic elements which are the symbol candidates as the image storage means (2).
The original image information stored in is cut out, it is determined whether or not it is a symbol, and the symbol code is stored in the symbol storage means (8) for the one determined to be a symbol. The character storage means (9) also cuts out the original image stored in the image storage means (2) from the information of the graphic elements of a certain size which are separated character candidates, and determines whether or not it is a character. If the character is determined to be a character, the character code is stored in the character storage means (10).

〔発明が解決しようとする問題点〕[Problems to be solved by the invention]

従来の図形認識装置は以上のように構成され,第2図に
示すような大図面(例えばA0やA1サイズ)に書かれた文
字や孤立シンボルを分離,認識するためにはデータ量が
膨大となるため,分割して処理しなければならず画面を
分割して文字やシンボルを認識するには,境界のところ
の文字やシンボルを分離するのに問題点があつた。ま
た,細線化に時間を要したり文字やシンボルを認識する
のに再度画像情報を切り出すため時間を要するという問
題点があつた。
The conventional figure recognition device is configured as described above, and a large amount of data is required for separating and recognizing characters and isolated symbols written on a large drawing (for example, A0 or A1 size) as shown in FIG. Therefore, in order to recognize the characters and symbols by dividing the screen, there is a problem in separating the characters and symbols at the boundary. In addition, there is a problem that it takes time to thin the lines and it takes time to cut out image information again in order to recognize characters and symbols.

この発明は,このような問題点を解決するためになされ
たもので,大図面から文字や孤立シンボルを圧縮した画
像データから分離し,それをもとに認識するため,細線
化や画像の切出しを行なわないで文字や孤立シンボルを
認識する図形認識装置を得ることを目的とするものであ
る。
The present invention has been made in order to solve such a problem. In order to separate a character or an isolated symbol from a large drawing from compressed image data and to recognize it based on that, thinning or cutting of an image is performed. It is an object of the present invention to obtain a figure recognition device for recognizing characters and isolated symbols without performing.

〔問題点を解決するための手段〕[Means for solving problems]

この発明にかかる図形認識装置は,大容量の画像データ
を圧縮して符号化する画像符号化手段と,この画像符号
化手段によって作成された符号化した画像情報を記憶す
る符号化画像記憶手段と、上記符号化した画像情報を所
定の行数分格納するバッファを有し、上記符号化画像記
憶手段によって記憶された符号化した画像情報を、上記
所定の行数毎に分割し1つ前の領域と最大シンボルの大
きさ分重複させたそれぞれの領域に区分して上記バッフ
ァに順次入力し、この入力した各領域毎の符号化した画
像情報から大きさが一定以下の図形の情報を分離する図
形情報分離手段と、この分離された図形の情報を記憶す
る図形情報記憶手段と、この図形情報記憶手段によって
記憶された図形の情報を入力して孤立シンボルを認識す
る図形情報認識手段と、上記図形情報記憶手段によって
記憶されている図形の情報のうち、上記図形情報認識手
段によって認識されたシンボル以外の情報を入力して文
字を認識する文字認識手段とを設けたものである。
A graphic recognition apparatus according to the present invention includes an image encoding unit that compresses and encodes a large amount of image data, and an encoded image storage unit that stores the encoded image information created by the image encoding unit. A coded image information stored in the coded image storage means is divided by the predetermined number of lines, and the buffer is provided for storing the coded image information for a predetermined number of lines. The area is divided into areas overlapping by the size of the maximum symbol, and the areas are sequentially input to the buffer, and the information of the graphic whose size is smaller than a certain size is separated from the input image information of each area. Graphic information separating means, graphic information storing means for storing information of the separated graphic, and graphic information recognizing hand for recognizing the isolated symbol by inputting the graphic information stored by the graphic information storing means. When, among the information of the graphic stored by said pattern information storage means, it is provided with a a character recognizing section for recognizing characters by entering information other than recognized symbols by said pattern information recognition means.

〔作用〕[Action]

この発明においては,図面中の孤立シンボルや文字を圧
縮した符号化データより画面を分割しながら分離抽出し
認識することにより,大きな図面に書かれた大容量のデ
ータを扱う場合でも,メモリが少なくて済み処理時間も
短縮できる。
According to the present invention, even if a large amount of data written in a large drawing is handled, the memory is reduced by separately extracting and recognizing the screen from the encoded data obtained by compressing isolated symbols or characters in the drawing. Processing time can be shortened.

〔発明の実施例〕Example of Invention

第1図はこの発明の一実施例を示す全体構成図であり,
(1)は図面を読取つて2値の画像情報を作成する画像
入力手段,(2)は画像入力手段(1)によつて作成さ
れた画像情報を記憶する画像記憶手段,(3)はこの画
像記憶手段(2)によつて記憶されている画像情報を入
力し,画像データを黒画素のラン(値1の連なり)の長
さとして符号化する画像符号化手段,(4)は画像符号
化手段(3)によつて作成された符号化された画像デー
タ(以降符号化データと呼ぶ)を記憶する符号化画像記
憶手段,(5)は上記符号化データを所定の行数分格納
するバッファを有し、符号化画像記憶手段(4)によっ
て記憶されている符号化データを、上記所定の行数毎に
分割し1つ前の領域と最大シンボルの大きさ分重複させ
たそれぞれの領域に区分して上記バッファに順次入力
し、この入力した各領域毎の符号化データからある大き
さ以下の連結領域を文字や孤立シンボルの候補となる図
形情報として取り出す図形情報分離手段、(6)は図形
情報分離手段(5)によつて作成された文字や孤立シン
ボルの候補となる符号化データを記憶する図形情報記憶
手段,(7)は図形情報記憶手段(6)によつて記憶さ
れている符号化データを入力し,各連結領域単位に孤立
したシンボルか否かを認識する図形情報認識手段,
(8)は図形情報認識手段(7)によつて認識されたシ
ンボルコードを記憶するシンボル記憶手段,(9)は図
形情報記憶手段(6)によつて記憶されている符号化デ
ータのうち孤立シンボルと認識されなかつたものに対
し,文字か否かを認識する文字認識手段,(10)は文字
認識手段(9)によつて認識された文字コードを記憶す
る文字記憶手段である。
FIG. 1 is an overall configuration diagram showing an embodiment of the present invention.
(1) is an image input means for reading a drawing to create binary image information, (2) is an image storage means for storing the image information created by the image input means (1), and (3) is this Image coding means for inputting the image information stored by the image storage means (2) and coding the image data as the length of a run of black pixels (a series of values 1), (4) is an image code Encoded image storage means for storing the encoded image data (hereinafter referred to as encoded data) created by the encoding means (3), and (5) stores the encoded data for a predetermined number of rows. Each area which has a buffer and which is obtained by dividing the coded data stored by the coded image storage means (4) by the predetermined number of rows and overlapped with the area immediately before by the size of the maximum symbol. It divides into and sequentially inputs into the above buffer, and each of this input Graphic information separating means for extracting a connected area of a certain size or less from the encoded data for each area as graphic information that is a candidate for a character or an isolated symbol, and (6) is a character created by the graphic information separating means (5). The graphic information storage means (7) for storing the coded data that is a candidate for an isolated symbol or an isolated symbol is input with the coded data stored by the graphic information storage means (6) and isolated in each connected area unit. Graphic information recognition means for recognizing whether it is a symbol,
(8) is a symbol storage means for storing the symbol code recognized by the graphic information recognition means (7), and (9) is isolated from the encoded data stored by the graphic information storage means (6). Character recognition means for recognizing whether or not a character is not recognized as a symbol, and (10) is a character storage means for storing the character code recognized by the character recognition means (9).

次に上記実施例の動作を第2図〜第9図を参照しながら
説明する。例えば第2図に示すような孤立したシンボル
や文字及び線を含む図面を画像入力手段(1)により2
値の画像データとして画像記憶手段(2)に記憶する。
画像符号化手段(3)は例えば第3図に示すように画像
データに対し,各行の黒画素のひとつながりの始点と長
さを算出し,符号化する。このようにして符号化された
データ(ランレングスデータと呼ぶ)のx軸方向又はy
軸方向のみのランレングスデータがとり出され,符号化
画像記憶手段(4)に記憶される。図形情報分離手段
(5)は第4図に示すように,符号化されたランレング
スデータをバツフアサイズいつぱいに読込む(例えばエ
リア(α)が読込まれたとする)。読込まれた各ランレ
ングスデータには第5図に示すように,ひとつながりの
黒画素(これをランと呼ぶ)毎にラベル(第5図(a)
(b)中の,,…)をつける。次の行のランにラベ
ルをつける時は,第5図(a)に示すように前の行のラ
ンと重なりがない場合はあらたなラベルをつけ,重なり
がある場合は前の行と同一のラベルを付ける。また第5
図(b)に示すように,前の行の2つ以上のランと重な
る場合は,重なる前の行のランのうち最少のラベルをつ
ける。このようにして各行のすべてのランに対してラベ
ル付けする。また,各ラベル付けられたラン(これを1
つのオブジエクトと呼ぶ)毎にx,y座標の最大値,最小
値やオブジエクトを構成するランの数,オブジエクトの
関係(例えばランとラベル付けられたオブジエクトと
ランとラベル付けられたオブジエクトは同一)等々の
情報を登録する。こうしてすべてのランに対し,ラベル
付けが終了すると第6図に示すように,オブジエクトの
関係を参照にしてラベルの値の最も小さいオブジエクト
に対し,x,yの最大,最小,ラベル数を修正し,1つの孤立
領域のランレングスの情報とする。出力は第7図に示す
ようにエリア(α)のランレングスデータからシンボル
の候補となるような一定サイズのもののランを孤立領域
毎にとり出す。この時,各孤立領域毎に出力されるラン
の情報のエリアを確保しておくことにより,エリア
(α)のランレングス情報を一度サーチするだけで各孤
立領域毎のランレングスがとり出せる。次に第8図に示
すようにエリア(β)のランレングスデータを読込む。
この時エリア(α)とはd行(d>シンボルの最大の大
きさ)オーバーラツプして読込む。前のエリアから続く
孤立領域のうち,はエリア(α)で既に出力されて
いるためエリア(β)では出力しない。また領域はエ
リア(α)では出力せず,エリア(β)で出力の判定を
行なう。あとは,エリア(α)と同様の方法で,各孤立
領域毎にラベル付けしてランレングスデータをとり出
す。こうして全体画面に対して,孤立シンボルや文字の
候補となる図形情報のランレングスデータをとり出し,
図形情報記憶手段(6)に記憶される。シンボル認識手
段(7)は第9図に示すように,シンボル候補となる各
孤立領域のランレングス情報に対し,縦,横の大きさ,
面積,内部の白領域数,円形度,点対称性等々の特徴量
を算出し,これとシンボル辞書に登録されている各々の
シンボルの特徴とを照合することによつてシンボルか否
か及びシンボルの場合はシンボルの種別を判定する。こ
うして判定されたシンボルの情報はシンボル記憶手段
(8)に記憶される。各孤立領域のランレングス情報に
対しシンボルと判定されなかつたものは,各孤立領域毎
に画像情報に復号化し,例えばパターンマツチング等々
により文字・数字か否かを照合する。判定された文字・
数字の情報は文字記憶装置に記憶される。
Next, the operation of the above embodiment will be described with reference to FIGS. For example, a drawing including isolated symbols, characters and lines as shown in FIG. 2 can be displayed by the image input means (1).
The value is stored in the image storage means (2) as image data.
For example, as shown in FIG. 3, the image encoding means (3) calculates the start point and the length of the connection of the black pixels in each row and encodes the image data. The x-axis direction or y of the data encoded in this way (called run-length data)
Run length data only in the axial direction is extracted and stored in the encoded image storage means (4). As shown in FIG. 4, the graphic information separating means (5) reads the encoded run-length data at a buffer size and time (for example, it is assumed that the area (α) is read). As shown in FIG. 5, each read run length data has a label (FIG. 5 (a)) for each connected black pixel (this is called a run).
Add (...) in (b). When labeling the run of the next line, a new label is given if there is no overlap with the run of the previous line as shown in Fig. 5 (a), and if there is overlap, it is the same as the previous line. Give a label. The fifth
As shown in FIG. 6B, when two or more runs in the previous row overlap, the minimum label is given to the runs in the previous row. Label all runs in each row in this way. Also, for each labeled run (1
The maximum and minimum values of x and y coordinates for each object, the number of runs that make up the object, the relationship between the objects (for example, the object labeled run and the object labeled run are the same), etc. Register the information of. When labeling is completed for all runs in this way, as shown in Fig. 6, the maximum and minimum of x and y, and the number of labels are corrected for the object with the smallest label value by referring to the object relationship. Therefore, the run length information of one isolated area is used. As shown in FIG. 7, the output takes out, for each isolated region, a run of a fixed size that becomes a symbol candidate from the run length data of area (α). At this time, by securing an area for the run information to be output for each isolated area, the run length for each isolated area can be obtained by searching the run length information of area (α) once. Next, as shown in FIG. 8, the run length data of the area (β) is read.
At this time, the area (α) is overlapped and read in d rows (d> maximum size of symbol). Among the isolated areas continuing from the previous area, since the area (α) has already been output, it is not output in the area (β). Further, the area is not output in the area (α), but the output is determined in the area (β). After that, run length data is extracted by labeling each isolated area in the same manner as in area (α). In this way, run length data of graphic information that is a candidate for isolated symbols and characters is extracted for the entire screen,
It is stored in the graphic information storage means (6). As shown in FIG. 9, the symbol recognition means (7) has vertical and horizontal sizes with respect to the run length information of each isolated area which is a symbol candidate.
Whether the symbol is a symbol or not can be determined by calculating the feature amount such as the area, the number of white areas in the interior, the circularity, the point symmetry, and the like, and comparing it with the feature of each symbol registered in the symbol dictionary. If, then the type of symbol is determined. Information of the symbols thus determined is stored in the symbol storage means (8). If the run length information of each isolated area is not determined to be a symbol, it is decoded into image information for each isolated area, and whether it is a character or a number is checked by pattern matching or the like. Judged character
Numerical information is stored in the character storage device.

〔発明の効果〕〔The invention's effect〕

以上のようにこの発明によれば、圧縮し符号化した画像
情報を、所定の行数毎に分割し1つ前の領域と最大シン
ボルの大きさ分重複させたそれぞれの領域に区分してバ
ッファに順次入力し、大きさが一定以下の図形の情報を
分離する図形情報分離手段と、この分離した図形の情報
を入力して孤立シンボルを認識する図形情報認識手段
と、上記分離した図形の情報のうち、上記図形情報認識
手段によって認識されたシンボル以外の情報を入力して
文字を認識する文字認識手段とを設けたので、大容量の
図面に対しても少ないメモリにより従来より短時間で孤
立シンボル及び文字を認識することが可能となる。
As described above, according to the present invention, the compressed and coded image information is divided into a predetermined number of rows and is divided into buffer areas by being divided into the previous area and the maximum symbol size. To the figure information separating means for separating the information of the figures whose size is less than a certain value, the figure information recognizing means for recognizing the isolated symbols by inputting the information of the separated figures, and the information of the separated figures. Among them, the character recognition means for recognizing characters by inputting information other than the symbols recognized by the graphic information recognition means is provided, so that even a large-capacity drawing can be isolated in a shorter time than before with a small memory. It becomes possible to recognize symbols and characters.

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

第1図はこの発明の一実施例である図形認識装置を示す
全体構成図,第2図は認識対象となる図面例を示す図,
第3図は符号化された画像データの説明図,第4図は大
画面の図面を分割して処理するための説明図,第5図
(a)(b)は各ランへのラベルの付け方を示す説明
図,第6図はラベル付けられたラン情報を孤立領域毎に
まとめるための説明図,第7図は孤立領域毎のランの出
力方法を示す説明図,第8図は大画面のオーバーラツプ
領域にある孤立領域の出力方法を示す説明図,第9図は
ランレングス情報からシンボルを認識するための説明
図,第10図は従来の図形認識装置の全体構成図である。 図において,(1)は画像入力手段,(2)は画像記憶
手段,(3)は画像符号化手段,(4)は符号化画像記
憶手段,(5)は図形情報分離手段,(6)は図形情報
記憶手段,(7)は図形情報認識手段,(8)はシンボ
ル記憶手段,(9)は文字認識手段,(10)は文字記憶
手段,(11)は画像処理手段,(12)は線情報記憶手
段,(13)は図形要素分離手段,(14)は図形要素記憶
手段,(15)はシンボル認識手段である。 図中,同一符号は同一または相当部分を示す。
FIG. 1 is an overall configuration diagram showing a figure recognition device according to an embodiment of the present invention, and FIG. 2 is a diagram showing an example of a drawing to be recognized,
FIG. 3 is an explanatory diagram of encoded image data, FIG. 4 is an explanatory diagram for dividing and processing a large screen drawing, and FIGS. 5 (a) and 5 (b) are labeling methods for each run. , FIG. 6 is an explanatory view for collecting labeled run information for each isolated area, FIG. 7 is an explanatory view showing a run output method for each isolated area, and FIG. 8 is a large screen. FIG. 9 is an explanatory diagram showing an output method of an isolated region in the overlap region, FIG. 9 is an explanatory diagram for recognizing symbols from run length information, and FIG. 10 is an overall configuration diagram of a conventional graphic recognition device. In the figure, (1) is image input means, (2) is image storage means, (3) is image coding means, (4) is coded image storage means, (5) is graphic information separation means, and (6). Is a graphic information storage means, (7) is a graphic information recognition means, (8) is a symbol storage means, (9) is a character recognition means, (10) is a character storage means, (11) is an image processing means, and (12). Is a line information storage means, (13) is a graphic element separation means, (14) is a graphic element storage means, and (15) is a symbol recognition means. In the figure, the same reference numerals indicate the same or corresponding parts.

───────────────────────────────────────────────────── フロントページの続き (56)参考文献 特開 昭57−197681(JP,A) 特開 昭60−258689(JP,A) 特開 昭62−190576(JP,A) 特開 昭62−206685(JP,A) 特開 昭58−106961(JP,A) 特開 昭61−169966(JP,A) 特開 昭62−160582(JP,A) ─────────────────────────────────────────────────── ─── Continuation of the front page (56) Reference JP-A-57-197681 (JP, A) JP-A-60-258689 (JP, A) JP-A-62-190576 (JP, A) JP-A-62- 206685 (JP, A) JP 58-106961 (JP, A) JP 61-169966 (JP, A) JP 62-160582 (JP, A)

Claims (3)

【特許請求の範囲】[Claims] 【請求項1】図面中の文字や孤立したシンボルを検出し
て認識する図形認識装置において、図面を読取って2値
の画像情報を作成する画像入力手段と、この画像入力手
段によって作成された画像情報を記憶する画像記憶手段
と、この画像記憶手段によって記憶された画像情報から
値が1の黒画素の情報を符号化してとり出す画像符号化
手段と、この画像符号化手段によって作成された符号化
した画像情報を記憶する符号化画像記憶手段と、上記符
号化した画像情報を所定の行数分格納するバッファを有
し、上記符号化画像記憶手段によって記憶された符号化
した画像情報を、上記所定の行数毎に分割し1つ前の領
域と最大シンボルの大きさ分重複させたそれぞれの領域
に区分して上記バッファに順次入力し、この入力した各
領域毎の符号化した画像情報から大きさが一定以下の図
形の情報を分離する図形情報分離手段と、この分離され
た図形の情報を記憶する図形情報記憶手段と、図形情報
記憶手段によって記憶された図形の情報を入力して孤立
シンボルを認識する図形情報認識手段と、図形情報記憶
手段によって記憶されている図形の情報のうち、図形情
報認識手段によって認識されたシンボル以外の情報を入
力して文字を認識する文字認識手段とを備えたことを特
徴とする図形認識装置。
1. A graphic recognition device for detecting and recognizing a character or an isolated symbol in a drawing, and an image input means for reading the drawing to create binary image information, and an image created by the image input means. An image storage means for storing information, an image coding means for coding and extracting information of a black pixel having a value of 1 from the image information stored by the image storage means, and a code created by this image coding means. Coded image storage means for storing the coded image information, and a buffer for storing the coded image information for a predetermined number of rows, the coded image information stored by the coded image storage means, The data is divided into the predetermined number of lines and divided into respective regions which overlap with the previous region by the size of the maximum symbol and are sequentially input to the buffer, and the input regions are encoded. A graphic information separating means for separating graphic information of a size smaller than a certain size from image information, a graphic information storing means for storing information of the separated graphic, and a graphic information stored by the graphic information storing means are inputted. Character recognition for recognizing a character by inputting information other than the symbol recognized by the graphic information recognizing means among the graphic information recognizing means for recognizing an isolated symbol and the graphic information stored by the graphic information storing means. A pattern recognition device comprising:
【請求項2】図形情報分離手段として、符号化された画
像情報から各孤立領域毎にラベルを付け、孤立領域の大
きさ、ラベル数の情報を求めて図形情報として一定の大
きさの孤立領域のみの符号化された画像情報を、孤立領
域毎に図形情報として出力することを特徴とする特許請
求の範囲第(1)項記載の図形認識装置。
2. The graphic information separating means attaches a label to each isolated area from the encoded image information, obtains the size of the isolated area and the information of the number of labels, and obtains the isolated area of a certain size as the graphic information. The graphic recognition device according to claim (1), characterized in that the coded image information of only one is output as graphic information for each isolated area.
【請求項3】図形情報認識手段として、各孤立要素毎に
分離された図形情報に対し、各孤立要素の大きさ、対称
性、円形度、面積、内部白領域数の特徴量を算出し、こ
れとシンボル辞書に登録されている各々のシンボルの特
徴とを照合することによって図形情報を認識するように
した特許請求の範囲第(1)項記載の図形認識装置。
3. The graphic information recognizing means calculates feature quantities of the size, symmetry, circularity, area, and number of internal white areas of each isolated element for the graphic information separated for each isolated element, The graphic recognition apparatus according to claim (1), wherein graphic information is recognized by collating this with the characteristics of each symbol registered in the symbol dictionary.
JP62238413A 1987-09-22 1987-09-22 Figure recognition device Expired - Fee Related JPH07104940B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP62238413A JPH07104940B2 (en) 1987-09-22 1987-09-22 Figure recognition device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP62238413A JPH07104940B2 (en) 1987-09-22 1987-09-22 Figure recognition device

Publications (2)

Publication Number Publication Date
JPS6481081A JPS6481081A (en) 1989-03-27
JPH07104940B2 true JPH07104940B2 (en) 1995-11-13

Family

ID=17029837

Family Applications (1)

Application Number Title Priority Date Filing Date
JP62238413A Expired - Fee Related JPH07104940B2 (en) 1987-09-22 1987-09-22 Figure recognition device

Country Status (1)

Country Link
JP (1) JPH07104940B2 (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0830725A (en) * 1994-07-20 1996-02-02 Canon Inc Device and method for processing image
JP3607433B2 (en) * 1996-10-04 2005-01-05 株式会社リコー Method and apparatus for extracting electrical symbols from construction drawings
JP6775624B2 (en) * 2019-03-01 2020-10-28 株式会社システムズナカシマ Image processing equipment, image processing system, image processing method, and program

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS57197681A (en) * 1982-05-24 1982-12-03 Hitachi Ltd Drawing recognizing device
JPS60258689A (en) * 1984-06-06 1985-12-20 Matsushita Electric Ind Co Ltd Image recognizing device

Also Published As

Publication number Publication date
JPS6481081A (en) 1989-03-27

Similar Documents

Publication Publication Date Title
CN109933756B (en) Image file transferring method, device and equipment based on OCR (optical character recognition), and readable storage medium
US5774580A (en) Document image processing method and system having function of determining body text region reading order
EP0542566A2 (en) Character recognition method and apparatus thereof
CN112861736B (en) Document table content identification and information extraction method based on image processing
CN112381057A (en) Handwritten character recognition method and device, storage medium and terminal
JP3062382B2 (en) Image processing apparatus and method
JPH07104940B2 (en) Figure recognition device
JPH0991371A (en) Character display device
JPS58197581A (en) Method and device for recognizing character and figure
JP2606560B2 (en) Document image storage device
JP2746345B2 (en) Post-processing method for character recognition
JPH0658689B2 (en) Line figure separation device
JP2571236B2 (en) Character cutout identification judgment method
JPH07325921A (en) Device and method for extracting connected area
JP2679098B2 (en) Encoding processing device for contour detection image
JP3720405B2 (en) Region identification apparatus and method
JP2683711B2 (en) How to recognize / correct character / symbol data
JP2000200323A (en) Online handwriting kanji recognizing device
JPH07220028A (en) Character recognition device
JP2974167B2 (en) Large Classification Recognition Method for Characters
JP2824704B2 (en) Character recognition device
JPH0573685A (en) Rader chart recognizing device
JPH04309185A (en) Drawing element separating method for automatic drawing input device
JPH0343879A (en) Character area separating system for character recognizing device
JPH0142029B2 (en)

Legal Events

Date Code Title Description
LAPS Cancellation because of no payment of annual fees