JP3070801B2 - Drawing management method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to and maps, facilities drawings, such as the piping system, mechanical drafting, architectural diagrams, drawings management how the printed circuit board mounting diagram, a variety of design drawings, etc. and manages the electronic information of
In respect, in particular, it relates to drawing management for converting raster graphic information image input to the vector graphic information.

[0002]

2. Description of the Related Art In recent years, paper maps and various kinds of design drawings have been widely used for electronic management. As a method of digitizing drawing information, there are a method of converting vector information into which coordinates of a figure are stored and a method of converting raster information into storing drawing information as a set of points.

[0003] When a drawing is managed by vector information, the information is described as line segment coordinates and related information can be attached, so that vectorized data can be directly applied to calculation processing and the like, and data can be corrected using a mouse or tablet. This is suitable for drawing management with frequent correction. However, it takes time and effort to perform initial data input, and a large amount of drawing information imposes a heavy burden on the operator.

On the other hand, when a drawing is managed by raster information, a digitized initial input can be easily obtained by taking a paper drawing from an image data reader. However, the rasterized data indicates that the pixel has a density value or 2
It is represented by a value (1, 0) and is not suitable for use other than display and printing. Further, even if the correction can be easily performed on the paper drawing, the direct correction using the raster information is troublesome and lacks accuracy.

[0005] Under such circumstances, a processing technique for converting raster information read from a paper drawing into vector information has recently been developed. Such examples include “Image processing (p.1146-1150)” and “Character input (p.1)” published in Information Processing Handbook (1989, published by Ohmsha).
151 to 1155) ". Here, a vectorization technique in which a digital line image or a label image is obtained from a gray-scale surface image or a binary surface line image, and a line segment description and a region description are described.

Further, the rasterized data includes noise due to shrinkage of the paper, dimensional deviation due to the angle at the time of reading, poor reproducibility of fine and dense parts, dirt and wrinkles, and the like.
These correction methods are described in, for example, “GMM Drawing Input Device”.
ADG-1010 ”(Hitachi Review; Vol. 73,
No. 10 (1991); 51-56) ".

[0007]

When line information of raster information is recognized and converted into vector information, erroneous recognition and erroneous conversion at the time of reading from a paper drawing as described above often occur. Correction work to improve the quality of figures is indispensable.

Conventionally, the operator alternately monitors the drawing and data of the converted vector graphic information, such as removal of unnecessary information and correction processing according to the graphic element, for example, for a straight line, near line connection or linearization. While going.

In this case, in a drawing in which drawing elements are mixed,
It is necessary to improve the reproducibility of the converted figure by repeating the necessary correction processing for each figure element for one drawing many times interactively. I have.

An object of the present invention is to convert raster graphic information in which a plurality of drawing elements are mixed into vector graphic information.
It is to provide a drawing management method for performing pattern processing.

[0011]

[0012]

According to the present invention, there are provided a character, a symbol,
The drawing management method for identifying the drawing element from raster graphic information in which a plurality of drawing elements such as a straight line, a curve, and a plane figure (closed figure) are mixed and creating vector graphic information by vectorization or encoding, For each area of raster graphic information divided by the ratio or combination of
It is characterized in that processing patterns indicating the contents and order of the conversion processing and correction processing for creating vector graphic information are defined in advance, and the processing is performed according to the processing patterns.

[0013]

According to the present invention, the area of raster graphic information is divided by drawing elements such as "table" mainly composed of ruled lines and "path" mainly composed of symbols and straight lines. The processing content and order are defined for each area. As a result, the subsequent conversion processing and correction processing are automatically performed.

According to the present invention, differences and ratios of drawing elements,
Alternatively, for each combination, drawing classifications such as “table”, “route”, and “map” and processing patterns suitable for the classifications are defined in advance based on the experience of the skilled person and the past performance, so that the Is specified, a processing pattern corresponding to each area is executed, and corrected vector graphic information is created. Thereby, the burden on the operator in the area division and the process definition is reduced, and the quality of the vector graphic information can be improved.

As described above, according to the present invention, even in the case of a drawing in which elements are mixed, optimal processing can be patterned in accordance with the difference or ratio of elements, and automation of conversion processing can be realized. In addition, since raster information read from a paper drawing can be easily converted into vector information, a drawing management device that can easily maintain and manage digitized drawings, such as creation and correction, can be provided.

[0016]

An embodiment of the present invention will be described below with reference to the drawings.

FIG. 2 is a basic configuration diagram of a drawing management apparatus to which the present invention is applied. The drawing management device is a central processing unit (CP
U) 1, file device 2, drawing output device (printer)
3, mouse 4, main memory 5, display device (C
RT) 6, a keyboard 7, etc., and a drawing reading device 8 is additionally provided.

In FIG. 1, the file device 2 stores drawing data. The drawing data includes raster data that stores all characters, lines, and figures as point information, vector data that stores lines and figures as coordinate information, and coded (font) data that encodes and stores characters and symbols. There is. Vector data and encoded data are provided with attribute data that expresses related information corresponding to the figure or code by characters or numerical values.

The display screen of the CRT 6 includes an icon area 61 for selecting functions and a drawing display area 62 as shown in FIG. In order to display the drawing data on the CRT 6, the “read drawing” of the function selection icon is
Is operated and the cursor is designated by the CU.

The CPU 1 designates "read drawing",
When a data type (raster data or vector data), a drawing name, a drawing number, or the like is input from the keyboard 7, the corresponding drawing data is retrieved from the file device 2 and temporarily stored in the main memory 5, and this drawing data is stored. C
Edit and display according to the display area (coordinates) of RT6.

The main memory 5 stores programs and data being processed in order for the CPU 1 to perform processing such as search and editing of screen data.

The CPU 1 also has a well-known graphic display function such as enlargement or reduction of the display screen at an arbitrary magnification, scrolling when the user wants to see outside the display area, and rotation of the screen by designating a function selection icon. Therefore, the operator can see the contents of the target drawing in detail by the screen display by these functions.

The drawing data stored in the file device 2 or the main memory 5 is printed by a printer 3 on paper or polyester film.

Next, raster data for reading and storing a paper drawing and a process of converting raster data into vector data or characters / symbols will be described with reference to FIGS.
This will be described below. FIG. 1 is a functional block diagram of a drawing management apparatus according to the present embodiment, which is realized by a CPU 1 and its software. FIG. 3 is a flowchart showing an outline of the processing of the drawing management apparatus.

At the start of the operation of the drawing management apparatus starting from drawing reading, the "drawing input" icon is selected.
This selection command is input to the drawing management control unit 113 via the operation input unit 111, and the drawing reading device 8 is operated by a “drawing input” command from the drawing management control unit 113 to the drawing input unit 112.

In step s101, the drawing reader 8 scans a paper (paper or film) drawing at a predetermined speed, and converts the drawing into colors and shades according to the shading and color of each point.
Digitization is performed to generate drawing information based on raster data.

In step s102, the drawing management control unit 1
13 inputs the raster drawing information, temporarily stores it in the main memory 5, edits and displays it in accordance with the drawing display area 62 by the drawing management display processing unit 117, and attaches a predetermined search key to the raster information to generate a file. It is stored in the raster information area 20 of the device 2. Search keys include drawing numbers, such as "Pipeline", "Construction Plan", "Water Ledger",
A drawing name such as “architectural drawing” or “mechanical drawing” is given.

In step s103, desired raster graphic information is read out by a search key and displayed on the screen. On this screen, the operator segments the processing area with the mouse,
Set the drawing classification for each area. The coordinates of the processing area and the drawing classification are stored in the main memory 5. As will be described later, the raster graphic information is divided and stored for each divided area.

The processing area and the drawing classification may be set in step s102, and may be stored in the raster information area 23 together with the search key. According to this, when converting a plurality of drawings, continuous processing can be performed only by designating the drawings that need to be converted.

The drawing classification is classified in advance by names such as "straight line", "character", "table", "route", and "map" according to the difference, ratio or combination of graphic elements. The processing contents and order of “vector conversion”, “character / symbol encoding (also referred to as character / symbol recognition)”, and “graphic correction” suitable for each classification are stored in the processing definition table 23 of the file device 2. Is stored. The drawing classification and the process definition can be arbitrarily set by the operator each time.

In step s104, first, the icon of "automatic conversion correction" is selected. Thereby, the drawing management control unit 113 refers to the processing definition corresponding to the set drawing classification for each processing area of the drawing.

In step s105, the control unit 113 executes the vector conversion processing unit 11 according to the referenced process definition.
4 (s106), processing of the character / symbol recognition processing unit 115 (s107), processing of the graphic correction processing unit 116 (s1)
08) is operated to encode or vectorize the raster graphic in the area and correct it. Then, when all the processing areas have been completed (s109), the automatic conversion processing ends.

The converted and corrected vector data describes its constituent coordinate points in two-dimensional coordinates (X _i , Y _i ) and stores it in the vector data area 22 of the file device 2. The encoded data converted from the character / symbol is stored in the encoded data area 21 together with the coordinate point information. It should be noted that region coordinates are given to the encoded data for each character row (column) so that the encoded data can be combined with the original drawing.

FIG. 4 is a flowchart showing an example of the conversion process of the vector graphic processing unit 114. First, an image of raster data is input from the memory 215 (s2
01). Assuming that this data form is a gray-scale (lumped) image, discontinuous parts are extracted by calculating the degree of boundary by a differential method or the like (s202). As a result, a gray line image is obtained.

Next, ridge point detection (s203) is performed to obtain a digital line image having a width of 1. Alternatively, the digital linear image is obtained by binarizing the gray-scale linear image into a binary linear image (s204), and thinning the image within a range where connectivity is not lost (s205). If the raster data is a gray line image or 2
In the case of the data form of the value linear image, the processing before s204 becomes unnecessary.

The digital line image is Hough
A line segment is described by conversion (s206) or line tracing (s207).

In the Hough transform, when the coordinates of each point on a digital line are (X ₁ , Y ₁ ), the parameters of all the straight lines passing through those points (for example, the length and direction of the perpendicular from the origin) are set to 2 A dimensional histogram is formed, and a straight line is estimated from the peak.

The line tracing is represented by a starting point coordinate on a digital line and a sequence of unit steps (chain code) in eight successive directions.

The individual vector data described by the line segment is
By the area description processing (s208), the line segments of the predetermined graphic area are combined, edited, and made into a database.

FIG. 5 is a flowchart showing the processing of the character / symbol recognition processing unit 115. First, when an image of raster data is input from the memory 215 (s301),
The background is separated from the distribution of the density histogram, and the four sides of the character, which is a two-dimensional plane figure, are extracted and preprocessed to align the size and position (s302).

Next, character pattern features are extracted by a horizontal feature for cutting the character pattern into strips or an outermost point list method based on the irregularities of the character pattern outline (s303).
The character pattern classified by the feature is identified as a specific character by a pattern matching method (s304) or a structural analysis method (s305), and is encoded into a code of a fixed font (s306). Finally, the origin coordinates of the character (symbol) are set as two-dimensional coordinate information, and are stored in a database together with the code (s307).

In the pattern matching method, an input pattern and a standard pattern of each character type based on the identification dictionary are superimposed,
The character type that provides the maximum matching is recognized as a character of the input pattern. In the feature analysis method, for example, it is determined whether a series from the top to the bottom of the horizontal feature matches an identification transition table created for each character type (recognition automan).

The graphic correction processing unit 116 includes a proximity line combining process for making the constituent points of a graphic which are close to each other within a specific distance into one graphic, a linearizing process for erasing constituent points having a specific width or more of a broken line, It has functions such as corner correction for correcting the corners of a vector closed figure to a fixed angle, and erasing minute figures for erasing minute vector figures smaller than a specific value, and performs processing by setting each specific value as a parameter.

FIG. 6 is an example of a screen in which the drawing name “water supply ledger” is searched and displayed from the raster data area 20 of the file device 2. The "water supply ledger" includes table information (ruled lines and character information) in which the user's address, name, construction history, and the like are entered, piping information indicating the piping system of the user's house, and map information indicating roads and house frames near the user's house. Consists of

The table information includes an area A [(x ₁ , y ₁ ),
(X ₂ , y ₂ )], and is composed of straight lines (ruled lines) and graphic elements of characters. The piping information includes the area B [(x ₃ , y ₃ ),
(X ₄ , y ₄ )] and is composed of graphic elements of straight lines and symbols. Piping information is a type of route information that is the same as a plant system diagram, an electrical network, a road network, a railway network, or the like. Map information, the region C _{[(x 5, y 5),} (x 6,
y ₆ )], and is formed mainly with graphic elements such as straight lines, curves, and surface figures.

FIG. 7 shows a process definition table 200 for the drawing classification of each area of the “water supply ledger”. To create the process definition table 200, first, the diagonal coordinates of the areas A, B, and C divided by the difference or combination of the drawing elements are set. Next, the processing contents of each area are defined in the processing order.

FIG. 8 is a conceptual diagram for explaining the area division of the raster graphic data of the “water supply ledger” before the conversion processing based on the area designation. The input raster data stored in the storage area 20 includes the entire area. From this data, the data of the area A, the area B, and the area C are divided and stored. Further, difference processing is performed between the entire area data and the data of the A, B, and C areas to create data of the area D. The data of the area D includes the entire outer frame, boundary information between the areas, and the like. For this area, the processing contents and the order according to the graphic elements are defined.

As described above, the target raster graphic is divided into regions in advance, and the conversion and correction processing is performed for each region. Therefore, unnecessary duplication processing is eliminated and the speed can be increased.

In the above example, drawing classification of each area is not always necessary. However, if the drawing classifications are classified according to the combination or ratio of drawing elements and the processing contents and order suitable for each drawing classification are defined in advance, the processing contents and order of each area specify the drawing classification. Can be defined only by

FIG. 9 shows such a processing definition table. In the process definition table 201, the optimal process content and order are defined in advance for each generalized drawing classification. Therefore, when the operator sets the processing area and its drawing classification, a table equivalent to that in FIG. 8 can be immediately generated.

In the present embodiment, the drawing classification and the processing pattern arbitrarily created by the operator are also registered so that they can be reused. The “table” and “route” in FIG. 9 are initially arbitrarily created. If the drawing classification is subdivided according to the parameters of the correction processing, the range of selection can be expanded according to the situation of the figure and the purpose of use.

After setting the processing area and creating the processing definition table, the drawing management control unit 113 starts the conversion processing from the area A. First, the character / symbol recognition processing unit 115 is operated to perform character recognition in the area A, encode each character to give coordinates, manage each character string (line), and store it in the storage area 21.

Next, the vector conversion processing unit 114 is operated to trace the raster information of the area A in a line, and a break point of each line is described by coordinates to be vectorized. Thereafter, the graphic correction processing unit 116 is operated to correct the vectorized data,
It is stored in the storage area 22.

The vector data of the area A has a drawing classification of "table", the ruled lines forming the table are relatively long straight lines and have few constituent points, and the start and end points of the line segments are connected by straight lines. With the line segment. In this embodiment, the correction is made in accordance with the characteristics of the table information.

FIG. 10 shows an example of processing for correcting the vectorized table data of the area A. FIG. 10A shows the processing pattern of this embodiment. FIG. 8B shows the result of performing the near-line coupling after the straight line correction regardless of the processing pattern of the present embodiment.

Since there is no divided straight line in the table information, the vectorized figures (H01, H04) before correction are obtained.
(301), which are close to each other (H02).
Also, since the table information is an artificial figure, the original drawing has no noise. Therefore, all the minute figures remaining after the correction of the proximity coupling can be regarded as noise, and are deleted. Since all lines between the start point and the end point of the line segment after the close coupling should be straight lines, the number of constituent points of one figure (line segment) is reduced, and the bent portion is corrected in a straight line (H03).

On the other hand, if the straight line correction is performed first as shown in FIG. 7B, each divided short line is linearized (H0).
5). After that, when a combining process is performed to form one figure (H0
6), portions (305, 306) that cannot be linearized occur. In addition, when the minute figure is erased before the near line coupling, even the useful divided short lines are erased, so that a closed table figure cannot be reproduced.

As described above, in the table information, it is effective to improve the reproducibility accuracy by performing near line coupling first and then performing straight line correction or minute figure elimination.

In the vector data of the area B, the drawing classification is “path” and the path information is an artificial figure.
Since a line segment is divided by a bent portion or a symbol of a route, a short line figure and a long line figure are mixed. Therefore, the path information requires a correction processing procedure that leaves a short line figure. Also, the route information is used for various network analyses, as is the case with the pipeline graphic of the present example, and the line graphic is not divided by a symbol other than the symbol (even when the symbol is not divided by the symbol). Yes).

FIG. 11 shows an example of processing for correcting the vectorized path data of the area B. FIG. 11A shows the processing pattern of the present embodiment, and FIG. 11B shows the processing pattern of the present embodiment. This is the case not depending on the pattern.

First, the character / symbol recognition processing unit 115 is operated to encode the symbols (a) and (b) which are apt to be mistaken for a figure, and then to perform vector graphic processing (K01).

Since there are many short lines in the path information (in this example, the pipeline information), first, as shown in FIG. 11A, the bent portions (402, 405) having a certain width or less are linearized by straight line correction. Line segments and bends that need to be left (403, 40
4,408,410) is evident (K02). Next, the segmented portions (401, 405, 407, 409) of the line segments are connected in close proximity to each other and are combined with other graphics to make sense as the route information (K03). This connecting portion becomes a new bent portion. The minute figure remaining after the proximity coupling is regarded as noise and is eliminated.

On the other hand, as shown in FIG. 13B, when the proximity line coupling is processed first, for example, 406 and 407 are coupled to form a bent portion having a certain width or less (K05). Is performed, the bent portions of 406 and 407 are removed. As a result, in FIG. 11A, the eight bent portions are reduced to four in FIG. 11B, and the bent shape of the conduit is omitted, so that the reproducibility is significantly reduced. .

As described above, in the route information, it is effective to correct the straight line first and then perform the near-line coupling.

Next, a conversion process of the map information of the area C is performed. Since the graphic in the area C in FIG. 7 does not include characters or symbols, the vector graphic processing and the graphic correction processing are directly performed. The map information is originally a figure based on a free curve of a natural object, and is mainly formed of a curved line, a curved line, and a closed figure, which are a collection of short line segments.

Accordingly, when there are many short straight lines, each straight line is straightened by straight line correction, and then the broken portions of the line segments are joined by proximity coupling. After that, corner correction for making a house frame, an intersection of roads, or the like a right angle is performed, and finally, a minute figure is deleted. When there are many curves, the straight line correction is not performed and the process is started from the proximity coupling.

According to the present embodiment, even for graphic information in which a plurality of graphic elements are mixed, the conversion and correction of a raster graphic into a vector graphic can be automatically performed only by setting the processing definition table. Therefore, if a paper drawing is input by the reading device 8, the conversion into vector data can be easily realized, and the maintenance of the drawing including the correction becomes extremely easy.

[0068]

According to the drawing management method of the present invention, the conversion from raster graphic information in which a plurality of drawing elements are mixed into vector graphic information can be patterned according to the ratio and combination of drawing elements. This has the effect of saving labor and improving the accuracy of converted data.

[0069]

[Brief description of the drawings]

FIG. 1 is a functional block diagram of a drawing management apparatus according to an embodiment of the present invention.

FIG. 2 is a hardware configuration diagram of a drawing management apparatus as one application example of the present invention.

FIG. 3 is a flowchart illustrating a drawing management method according to an embodiment of the present invention.

FIG. 4 is a flowchart illustrating a conversion process from raster to vector data.

FIG. 5 is a flowchart illustrating a conversion process from raster characters to encoded data.

FIG. 6 is a screen display example of a “water supply ledger” drawing showing its area divisions.

FIG. 7 is a process definition table for each area of the “water supply ledger” drawing.

FIG. 8 is a conceptual diagram illustrating divided storage of raster data for each area.

FIG. 9 is a process definition table defined by a drawing classification.

FIGS. 10A and 10B are screen display examples for each processing step in correcting “table” information, where FIG. 10A shows a case according to the processing procedure of the present embodiment, and FIG.

FIGS. 11A and 11B are screen display examples for each processing step in correcting “route” information, where FIG. 11A shows a case where the processing procedure according to the present embodiment is used and FIG.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Central processing unit (CPU) 2 ... File device, 3 ... Drawing output device, 4 ... Mouse, 5 ... Main memory, 6 ... Display, 7 ... Keyboard, 8 ... Drawing reading device, 20 ...
Raster data storage area, 21: character / symbol data storage area, 22: vector data storage area, 23: processing definition table storage area, 61: operation icon area, 62
... Drawing display area, 111 ... Operation input unit, 112 ... Drawing input unit, 113 ... Drawing management control unit, 114 ... Vector graphic processing unit, 115 ... Graphic recognition processing unit, 116 ... Graphic correction processing unit, 117 ... Drawing management Display processing unit, 200, 201
... Process definition table.

────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-5-233811 (JP, A) JP-A-3-31970 (JP, A) JP-A-5-20493 (JP, A) (58) Field (Int.Cl. ⁷ , DB name) G06T 5/00

Claims (3)

(57) [Claims] 1. A raster graphic in which a plurality of drawing elements are mixed.
A drawing management method that converts information into vector graphic information
The raster graphic information by the ratio or combination of the drawing elements.
And split it into one or more areas.
Define the contents and order of conversion and correction
Raster of each area according to the definition of conversion processing and correction processing
When converting the graphic information into the vector graphic information, if the raster graphic information is a “table” mainly composed of straight lines (for example, ruled lines), the correction processing is performed first for proximity line combination, and then for straight line correction and / or minute correction. A drawing management method characterized by performing graphic erasure. 2. A raster graphic in which a plurality of drawing elements are mixed.
A drawing management method that converts information into vector graphic information
The raster graphic information by the ratio or combination of the drawing elements.
And split it into one or more areas.
Define the contents and order of conversion and correction
Raster of each area according to the definition of conversion processing and correction processing
When converting graphic information into vector graphic information, when the raster graphic information is a “path” mainly including a symbol and a straight line,
A drawing management method, characterized in that, after conversion processing of symbol encoding and vectorization of a straight line, the correction processing first performs straight line correction, and then performs near line combination. 3. Characters, symbols, straight lines, curves, surface figures (closed figure)
Raster graphic information in which multiple drawing elements such as
Identify the drawing elements from the above, vectorization and encoding conversion,
Drawing management to create vector graphic information by performing correction processing
In the method, the drawing classification classified by the difference or combination of the drawing elements
The contents and order of the conversion process and the correction process are
The stored processing patterns are stored in advance, and the drawing classification includes “table” combining characters and ruled lines,
It has a "path" that combines a symbol and a straight line, and the "table"
The processing pattern defines encoding of characters and vectorization of ruled lines, then proximity line joining, and then straight line correction. The processing pattern of the “path” is encoding of symbols and vectorization of straight lines, and then linear correction, to define the subsequent proximity line coupling, the raster graphic information on the difference and combination of the drawing elements
Before dividing into more areas and corresponding to each area
Designated drawing classification and falls under the specified drawing classification
And the processing pattern defined as above
Accordingly , a drawing management method , wherein the conversion processing and the correction processing are performed .