JPH08202883A - Method for recognizing paint-out graphic in automatic drawing input device - Google Patents
Method for recognizing paint-out graphic in automatic drawing input deviceInfo
- Publication number
- JPH08202883A JPH08202883A JP7009361A JP936195A JPH08202883A JP H08202883 A JPH08202883 A JP H08202883A JP 7009361 A JP7009361 A JP 7009361A JP 936195 A JP936195 A JP 936195A JP H08202883 A JPH08202883 A JP H08202883A
- Authority
- JP
- Japan
- Prior art keywords
- vector
- executed
- data
- processing
- satisfied
- 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
Landscapes
- Image Analysis (AREA)
Abstract
Description
【0001】[0001]
【産業上の利用分野】この発明は、図面自動入力装置に
おける塗り潰し図形の認識方法に関する。BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for recognizing a filled figure in an automatic drawing input device.
【0002】[0002]
【従来の技術】図面自動入力装置は、図3に示す処理手
順になる。入力対象となる図面はイメージスキャナ等に
よって2値画像化する入力処理を行い(ステップS
1)、該2値画像に対して連続した画素のつながりとし
ての輪郭ベクトル群の抽出と、その情報圧縮になる輪郭
ベクトル処理を行い(ステップS2)、閉じた輪郭ベク
トル群から図形線分の芯線ベクトル群を抽出する芯線ベ
クトル処理(なお、芯線ベクトルは輪郭ベクトルからペ
アベクトルを検索し、その中心に生成してもよい。)を
行い(ステップS3)、これらベクトル群が閉じた孤立
要素について文字候補と図形候補(シンボル候補とその
他の線分候補)に分類する図面要素の分離を行う(ステ
ップS4)。この分類条件は独立要素(ベクトルの集合
体)の領域四角形の大きさにより設定したしきい値より
小さいものは文字候補、それ以外は図形候補とする。ま
た、文字候補はその集合の中から文字同志の横並び/縦
並び関係から複数の文字列に分類する。さらに、1つの
文字列候補の中から文字の大きさと文字間隔を基に1文
字候補を抽出する。抽出した1文字候補はベクトルの集
合体になる。2. Description of the Related Art A drawing automatic input device has a processing procedure shown in FIG. The drawing to be input is subjected to an input process of converting it into a binary image by an image scanner or the like (step S
1) Extracting a contour vector group as a series of continuous pixels from the binary image and performing a contour vector process for compressing the information (step S2), and extracting the core line of the graphic line segment from the closed contour vector group. Skeleton vector processing for extracting the vector group (note that the core vector may be searched for a pair vector from the contour vector and generated at the center) (step S3), and character processing is performed on the isolated element in which these vector groups are closed. Drawing elements to be classified into candidates and figure candidates (symbol candidates and other line segment candidates) are separated (step S4). This classification condition is a character candidate if it is smaller than the threshold value set by the size of the area quadrangle of the independent element (collection of vectors), and a figure candidate otherwise. In addition, the character candidates are classified into a plurality of character strings from the set based on the horizontal / vertical arrangement relationship between the characters. Furthermore, one character candidate is extracted from one character string candidate based on the character size and character spacing. The extracted one-character candidates become a set of vectors.
【0003】図形候補はその領域四角形の包含関係を基
にグループの細分化を施す。グループ化した図形候補単
位に属するベクトルを構造解析し、シンボル候補と線分
候補を抽出する。抽出した各シンボル候補及び線分候補
もまたベクトルの集合体になる。分離された文字候補は
各候補別にパターン比較のための辞書と照合することで
最も類似する文字を抽出する文字認識をし(ステップS
5)、同様にシンボル候補について最も類似するシンボ
ルを認識し(ステップS6)、これら認識結果は文字,
シンボルのコードとして抽出される。線分候補について
は整形を行う線分処理を行う(ステップS7)。The figure candidates are subdivided into groups based on the inclusion relation of the area rectangles. Structural analysis is performed on the vectors belonging to the grouped graphic candidate units, and symbol candidates and line segment candidates are extracted. Each extracted symbol candidate and line segment candidate also becomes a set of vectors. The separated character candidates are collated with a dictionary for pattern comparison for each candidate to perform character recognition for extracting the most similar character (step S
5) Similarly, the most similar symbol is recognized for the symbol candidates (step S6).
It is extracted as the code of the symbol. Line segment processing for shaping the line segment candidates is performed (step S7).
【0004】ステップS5〜S7で認識,処理された文
字及び図形データは結合・編集され(ステップS8)、
表示や印刷されるほかに図形データ処理を行うためのC
AD等のホストシステムへのデータ転送がなされる。こ
こで、輪郭ベクトルの抽出には図面の背景になる白画素
と図形・文字の黒画素との境界部分を追跡した黒画素列
を求め、この黒画素列が直線近似できるときは複数の画
素列を1つのベクトルデータとする情報圧縮が行われ
る。The character and graphic data recognized and processed in steps S5 to S7 are combined and edited (step S8),
C for displaying and printing as well as graphic data processing
Data transfer to a host system such as AD is performed. Here, in order to extract the contour vector, a black pixel row that traces the boundary between the white pixel that is the background of the drawing and the black pixel of the figure / character is obtained, and if this black pixel row can be linearly approximated, multiple pixel rows Is performed as one vector data.
【0005】[0005]
【発明が解決しようとする課題】前述した図面自動入力
装置を使用して図4に示す塗り潰し図形の認識を行って
その図形データをベクトル化すると、塗り潰し図形でな
い場合には図5に示すように外周輪郭ベクトルA、内周
輪郭ベクトルB、芯線ベクトルCと図4における中央部
分の丸印の外周輪郭ベクトルDが生成されるけれども、
塗り潰し図形の場合には、図5の外周輪郭ベクトルAの
みしか生成されないために、芯線ベクトルCは生成され
なくなる。このため、図形データとしては図6に示すよ
うになって元の図形とは異なった認識結果になってしま
う問題がある。When the filled figure shown in FIG. 4 is recognized and the figure data is vectorized by using the above-described drawing automatic input device, when the figure data is not filled, as shown in FIG. Although the outer peripheral contour vector A, the inner peripheral contour vector B, the core line vector C, and the outer peripheral contour vector D of the circle in the central portion in FIG. 4 are generated,
In the case of a filled figure, since only the outer peripheral contour vector A of FIG. 5 is generated, the core line vector C is not generated. Therefore, there is a problem that the figure data becomes as shown in FIG. 6 and the recognition result is different from the original figure.
【0006】この発明は上記の事情に鑑みてなされたも
ので、正確な図形認識ができるようにした図面自動入力
装置における塗り潰し図形の認識方法を提供することを
目的とする。The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a method of recognizing a filled figure in an automatic drawing input device which enables accurate figure recognition.
【0007】[0007]
【課題を解決するための手段および作用】この発明は、
上記の目的を達成するために、第1発明は、図面を2値
化画像として読取り、この2値化画像の輪郭ベクトルを
抽出し、この輪郭ベクトルから芯線ベクトルを生成して
図形・文字を認識する図面自動入力装置において、前記
輪郭ベクトルのうち外周輪郭ベクトルだけで、かつ、線
幅が一定以上であるかを判断したのち、その判断の条件
が満足したならベクトルデータは塗り潰しとして図形変
換処理を行うことを特徴とするものである。Means and Actions for Solving the Problems
In order to achieve the above object, the first invention reads a drawing as a binarized image, extracts a contour vector of the binarized image, generates a core vector from the contour vector, and recognizes a figure / character. In the drawing automatic input device, after determining whether only the outer peripheral contour vector among the contour vectors and the line width is equal to or more than a certain value, if the condition for the determination is satisfied, the vector data is subjected to graphic conversion processing as painting. It is characterized by performing.
【0008】第2発明は、前記判断の結果、条件が満足
しなかったときにはベクトルデータは一括ポリゴン、ポ
リライン処理を行うことを特徴とするものである。A second aspect of the present invention is characterized in that, when the result of the determination is that the conditions are not satisfied, the vector data is subjected to collective polygon and polyline processing.
【0009】[0009]
【実施例】以下この発明の一実施例を図面に基づいて説
明する。図1はこの発明の一実施例の処理フローチャー
トで、図1において、まず、図面を図3で述べたと同様
にステップS11でイメージスキャナ等により2値画像
化する入力処理を行う。この2値画像化されたデータに
対して連続した画素のつながりとしての輪郭ベクトル群
の抽出とその情報圧縮になる輪郭ベクトル処理を行って
ステップS12で画像データをベクトルデータに変換す
る。変換されたベクトルデータからステップS13で外
周輪郭ベクトルだけでベクトル間距離が1mm以上の線
幅があるかどうかを判断し、判断の結果、条件を満足し
たときにはステップS14の処理を行う。ステップS1
4の処理は塗り潰しとなるベクトルを選択する工程で、
このステップS14で選択されたデータは塗り潰し図形
変換をステップS15で行う。このステップS15で行
われた図形変換データを基にステップS17で編集修正
を行う。なお、ステップS13で条件を満足しなかった
ときには、ステップS16の処理で一括ポリゴン、ポリ
ライン認識を行い、この認識の結果はステップS17に
供給されて編集、修正処理される。DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a flow chart of an embodiment of the present invention. In FIG. 1, first, in the same manner as described with reference to FIG. 3, in step S11, an input process for converting a binary image by an image scanner or the like is performed. Extraction of a contour vector group as a continuous pixel connection and contour vector processing for information compression are performed on the binary imaged data, and the image data is converted into vector data in step S12. In step S13, it is determined from the converted vector data whether or not there is a line width with an inter-vector distance of 1 mm or more using only the outer contour vector. If the result of the determination is that the condition is satisfied, step S14 is performed. Step S1
The process of 4 is a process of selecting a vector to be filled,
The data selected in step S14 is converted into a filled figure in step S15. Editing and correction are performed in step S17 based on the graphic conversion data obtained in step S15. When the conditions are not satisfied in step S13, collective polygon and polyline recognition is performed in step S16, and the result of this recognition is supplied to step S17 for editing and correction processing.
【0010】上記のようにして塗り潰し図形を処理すれ
ば図2に示すように中央部分には塗り潰した丸印の図形
が、また、外側の四角形はポリゴン認識された図形デー
タが得られるので、正確な図形認識ができるようにな
る。When the filled figure is processed as described above, a filled circle figure is obtained in the central portion and polygon data is obtained for the outer quadrangle as shown in FIG. It becomes possible to recognize various figures.
【0011】[0011]
【発明の効果】以上述べたように、この発明によれば、
図面自動入力装置を使用して塗り潰し図形を認識しても
正確な図形を認識できるようになる利点がある。As described above, according to the present invention,
Even if a filled figure is recognized using the drawing automatic input device, there is an advantage that an accurate figure can be recognized.
【図1】この発明の一実施例を示すフローチャート。FIG. 1 is a flowchart showing an embodiment of the present invention.
【図2】実施例により得られた図形データ説明図。FIG. 2 is an explanatory diagram of graphic data obtained by the embodiment.
【図3】従来の図面自動入力装置の処理フローチャー
ト。FIG. 3 is a processing flowchart of a conventional drawing automatic input device.
【図4】塗り潰し図形説明図。FIG. 4 is an explanatory diagram of a filled figure.
【図5】ベクトルデータ説明図。FIG. 5 is an explanatory diagram of vector data.
【図6】従来例による塗り潰し図形の認識結果の説明
図。FIG. 6 is an explanatory diagram of a recognition result of a filled figure according to a conventional example.
S11…入力処理ステップ S12…ベクトルデータ変換ステップ S13…判断ステップ S14…塗り潰しデータ選択ステップ S15…塗り潰し図形変換ステップ S16…ポリゴン、ポリライン認識ステップ S17…編集、修正ステップ S11 ... Input processing step S12 ... Vector data conversion step S13 ... Judgment step S14 ... Fill data selection step S15 ... Fill figure conversion step S16 ... Polygon / polyline recognition step S17 ... Editing, correction step
Claims (2)
値化画像の輪郭ベクトルを抽出し、この輪郭ベクトルか
ら芯線ベクトルを生成して図形・文字を認識する図面自
動入力装置において、 前記輪郭ベクトルのうち外周輪郭ベクトルだけで、か
つ、線幅が一定以上であるかを判断したのち、その判断
の条件が満足したならベクトルデータは塗り潰しとして
図形変換処理を行うことを特徴とする図面自動入力装置
における塗り潰し図形認識方法。1. A drawing is read as a binarized image,
An automatic drawing input device that extracts a contour vector of a binarized image, generates a core vector from the contour vector, and recognizes a figure / character. The method for recognizing a filled figure in an automatic drawing input device, characterized in that, if the condition for the decision is satisfied, the vector data is subjected to a figure conversion process.
ときにはベクトルデータは一括ポリゴン、ポリライン処
理を行うことを特徴とする請求項1記載の図面自動入力
装置における塗り潰し図形認識方法。2. The method for recognizing a filled pattern in an automatic drawing input device according to claim 1, wherein when the result of the judgment is that the condition is not satisfied, the vector data is subjected to collective polygon and polyline processing.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP7009361A JPH08202883A (en) | 1995-01-25 | 1995-01-25 | Method for recognizing paint-out graphic in automatic drawing input device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP7009361A JPH08202883A (en) | 1995-01-25 | 1995-01-25 | Method for recognizing paint-out graphic in automatic drawing input device |
Publications (1)
Publication Number | Publication Date |
---|---|
JPH08202883A true JPH08202883A (en) | 1996-08-09 |
Family
ID=11718344
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
JP7009361A Pending JPH08202883A (en) | 1995-01-25 | 1995-01-25 | Method for recognizing paint-out graphic in automatic drawing input device |
Country Status (1)
Country | Link |
---|---|
JP (1) | JPH08202883A (en) |
-
1995
- 1995-01-25 JP JP7009361A patent/JPH08202883A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US4757551A (en) | Character recognition method and system capable of recognizing slant characters | |
CN101855640B (en) | Method for image analysis, especially for mobile wireless device | |
US4524454A (en) | Method of assigning direction code to boundary picture element in character recognition system | |
US20020154815A1 (en) | Character recognition device and a method therefore | |
JP3851742B2 (en) | Form processing method and apparatus | |
EP0933719A2 (en) | Image processing method and apparatus | |
JP2000207489A (en) | Character extracting method and device and record medium | |
CN114581932A (en) | Picture table line extraction model construction method and picture table extraction method | |
JPS60153574A (en) | Character reading system | |
Jia et al. | Grayscale-projection based optimal character segmentation for camera-captured faint text recognition | |
JPH08202883A (en) | Method for recognizing paint-out graphic in automatic drawing input device | |
JP3276555B2 (en) | Format recognition device and character reader | |
Chiou et al. | Recognition of Chinese business cards | |
JP2003196592A (en) | Program for processing image, and image processor | |
JP4194309B2 (en) | Document direction estimation method and document direction estimation program | |
JPH0877293A (en) | Character recognition device and generating method for dictionary for character recognition | |
JPH11120291A (en) | Pattern recognition system | |
JP2894111B2 (en) | Comprehensive judgment method of recognition result in optical type character recognition device | |
JP3013442B2 (en) | Automatic drawing input device | |
KR970002424B1 (en) | Candidate character classifying method | |
JP2571236B2 (en) | Character cutout identification judgment method | |
JP2674475B2 (en) | Character reader | |
KR100241447B1 (en) | English writing/number recognition method using outline information | |
JPH04222057A (en) | Method for recognizing character/graphic element for automatic drawing input device | |
JPS63269267A (en) | Character recognizing device |