JPH043274A - Vector piece number reduction processing system - Google Patents

Abstract

PURPOSE:To execute the processing at a high speed by setting the upper limit number of the final vector piece number obtained by a vector piece number reduction processing, and executing repeatedly the vector piece number reduction processing until the vector piece number obtained by the vector piece number reduction processing becomes below the designated upper limit number. CONSTITUTION:Under the control of a main processor 17, as a parameter at the time of converting image data to a vector, the upper limit number of the vector piece number outputted finally by a vector piece number reduction processing is set 184, and a vector conversion processing 14 for converting the image data to vector data is executed. Subsequently, whether the vector piece number obtained by the vector conversion processing 14 or the post- processing 16 is below the designated upper limit number or not is decided, and when the vector piece number is larger than the designated upper limit number, a prescribed vector piece number reduction processing for decreasing the vector piece number is executed until the vector piece number becomes below the upper limit number. In such a way, the image data can be converted automatically and quickly to the vector data having the prescribed output vector number.

Description

[Detailed Description of the Invention] [Summary] Regarding vector number reduction processing that reduces the number of vectors obtained by vectorizing image data, the present invention automatically and quickly converts image data into vector data having a predetermined number of output vectors. In a vector number reduction processing method that reduces the number of vectors obtained by vectorization processing that converts image data read from a drawing into vector data, the vector number reduction processing is used as a parameter when vectorizing image data. Set the upper limit for the number of vectors that will be finally output, execute the vectorization process to convert the image data to vector data, and check if the number of vectors obtained by the vectorization process or post-processing is less than or equal to the specified upper limit. If the number of vectors is larger than the specified upper limit number, a predetermined vector number reduction process is executed to reduce the number of vectors until the number of vectors becomes equal to or less than the upper limit number.

[Industrial application field]

The present invention relates to a vector number reduction processing method that reduces the number of vectors obtained by vectorization processing that converts image data read from a drawing into vector data.

(Prior art) CA D (computer A 1ded De
sign, computer-aided design) system, CA
M (computer aided manufac
2. Description of the Related Art In drawing input systems such as tuaring (computer-aided production) systems and computer mapping systems, automatic input systems are utilized to improve efficiency and accuracy when inputting drawings.

In this drawing automatic input system, the input image data is subjected to pre-processing such as binarization processing, image quality improvement processing, line thinning processing, etc., and then drawing recognition processing is performed to recognize the content of the drawing.

In drawing recognition processing, thinned image data is separated into line segments, characters, and symbols, and then vectorization processing is performed on the line segments, and vectorization processing is performed on the characters and symbols.
Respective recognition processing is performed.

This drawing recognition process is complex and the amount of data resulting from the process is enormous, so it is performed by a large or dedicated computer and requires a large memory capacity.

However, in addition to precisely reading input drawings, automatic drawing input systems can often be used to read schematic diagrams showing the rough contents of input drawings, such as background drawings and layout diagrams.

Such a schematic diagram requires less data, so
Drawing input processing can also be performed using a small computer such as a so-called office computer.

FIG. 4 shows a conventional image data vector number reduction processing method for obtaining a schematic diagram of such an input drawing.

In FIG. 4, an image scanner 22 reads figures (including characters) on a drawing 21 and sends the image data to a preprocessing section 23.

The preprocessing unit 23 performs a binarization process to binarize image data, an image quality selection process to remove various noises from the binarized image data, a process to extract features of the image data, and a process for each figure whose image quality has been changed. Line thinning processing is performed to narrow the line width and find the center line.

The vectorization processing unit 24 performs vectorization processing to convert the image data of each figure thinned by the preprocessing unit 23 into vector data. A typical example of vectorization processing is
This is a process of approximating a figure with polygonal lines and converting it into a series of coordinate data at both ends of each polygonal line.

The separation processing unit 25 performs a process of separating vector data into character, symbol, and line segment vector data using various known methods. Characters can be separated, for example, by the feature of isolated vector data sequences consisting of a number of short line segments within a small frame area. Also,
Symbols can be separated, for example, by matching with templates for each symbol.

Through this separation process, the number of characters, the number of symbols, and the number of vectors of all line segments are detected.

Since the number of line segments in a figure is large, the amount of vector data obtained in this way is extremely large.

However, since the schematic diagram does not require data regarding the fine structure of each figure, the post-processing unit 26 converts the highly accurate vector data obtained by the vectorization processing unit 24 into hector data of the schematic diagram, i.e. Vector number reduction processing is performed to reduce the number of vectors.

As a post-processing to reduce the number of vectors performed in the post-processing unit 26, for example, the linearization method (1) shown in FIG. 5(a) is used.
There is also a linearization method (2) shown in FIG.

Linearization method (1) is a method of converting two vectors a and b into one vector a' if the angle θ between them is close to 180 degrees, as shown in Figure (a). be.

This reduces two vectors to one vector. As the parameter θ is made smaller than 180 degrees, the graphic accuracy decreases, but the amount of decrease in the number of vectors increases.

Linearization method (2), as shown in the same figure (b), shows that two parallel lines sandwiching three consecutive vectors a, b, and C have a width of 1 and 2, and the width of the three vectors is 1 and 2. When the distance l from the starting point to the ending point is sufficiently small compared to the three vectors a,
This is a method of converting b and C into one straight line a' connecting their starting and ending points. This reduces the three vectors to one. As the value of the parameter h/l increases, the graphic accuracy decreases, but the amount of decrease in the number of vectors increases.

On the other hand, in small computers, due to memory capacity,
An upper limit is set for the number of vectors to be output.

Therefore, the operator should check the θ of the linearization method (1) and the linearization method (
The value of each parameter for the vector number reduction process such as h/I in 2) is set appropriately, and the post-processing section 26 executes the post-processing.

When the number of vectors after post-processing is greater than the specified upper limit number, the values of each parameter such as θ and h/1 are reset and post-processing is executed. Thereafter, the parameter values are corrected and post-processing is repeatedly executed until the number of vectors becomes less than the specified upper limit number.

In the manner described above, vector data of a figure that has been post-processed so that the number of vectors is equal to or less than a predetermined number of vectors is obtained, and is sent to a utilization device such as a host computer (not shown).

[Problem to be solved by the invention]

When reducing the number of vectors in the conventional image data vectorization method, as described above, the values of each parameter for vector number reduction processing are corrected until the number of vectors becomes less than the specified upper limit number. Post-processing to reduce the amount of damage was repeatedly executed.

Each time post-processing is repeated in this way, the operator must modify the parameters of each vector number reduction processing method, which requires a lot of time in the vector reduction post-processing section and places a heavy burden on the operator. There was a problem.

Furthermore, since parameter values are corrected by the operator through trial and error, it takes time to set them up and it is difficult to set them accurately, so post-processing is likely to be repeated many times, making post-processing more efficient. The problem was that it was difficult to

SUMMARY OF THE INVENTION An object of the present invention is to provide an improved image data vectorization method that can automatically and quickly convert image data into vector data having a predetermined number of output hectors.

[Means to solve the problem]

The means adopted by the present invention to solve the above-mentioned problems are as follows:
This will be explained with reference to the principle diagram shown in FIG. FIG. 1 shows a processing flowchart of the method for reducing the number of vectors according to the present invention.

In step Sl, the upper limit of the number of vectors to be finally output by the vector number reduction process is set as a parameter when vectorizing the image data.

In step S2, vectorization processing for converting image data into vector data is executed.

In step S3, it is determined whether the number of vectors obtained by vectorization processing or post-processing is less than or equal to a specified upper limit number.

In step S4, when the number of vectors is larger than the specified upper limit number, a vector number reduction process is executed to reduce the number of vectors according to a predetermined vector number reduction method. A specific method for this vector number reduction process will be explained in the next section.

[Effect]

The operation of the present invention will be explained according to its processing steps with reference to FIG. FIG. 2 shows an example of the vector number reduction processing method used in the present invention. It is assumed that the image data read from the drawing by an image scanner or the like has been subjected to line thinning processing by preprocessing (not shown).

(1) Step S1 Various parameters necessary for the vectorization process are set, and as one of the parameters, an upper limit number of vectors to be finally output by the vectorization process is set.

The upper limit of the number of vectors is determined by the required drawing reading accuracy, the memory capacity of the drawing input system, and the like.

(2) Step S2 Vectorization processing for converting the thinned image data into vector data is performed in the same manner as in the past. At this stage, the vector number reduction process has not been executed, so the number of vectors is an extremely large value.

(3) Step S.

It is determined whether the number of vectors obtained by the vectorization process is less than or equal to a specified upper limit number.

(4) Step S4 When the number of vectors is larger than the specified upper limit number, post-processing is performed to reduce the number of vectors according to a predetermined vector number reduction processing method.

FIG. 2 shows an example of the vector number reduction processing method used in the present invention.

FIG. 2(a) shows the same linearization method (1) as explained in FIG. 5(a). As shown in the figure, when there are two vectors a and b, if the angle θ between them is close to 180 degrees, then 1
This is a method of converting into a book vector a'. This results in
Two vectors are reduced to one vector. As the parameter θ is made smaller than 180 degrees, the graphic accuracy decreases, but the amount of decrease in the number of vectors increases.

FIG. 2 0)) is the same linearization method (2) as explained in FIG. 5 Q:1). As shown in the figure, the widths of two parallel lines k and 2 that sandwich three consecutive vectors a, b, and C are sufficiently smaller than the distance 1 from the starting point to the ending point of the three vectors. This is a method of converting three hectares a, b, and C into one straight line a' connecting the starting point and the ending point.

This reduces the three vectors to one.

As the value of the parameter h/l increases, the graphic accuracy decreases, but the amount of decrease in the number of vectors increases.

FIG. 2(c) shows the short vector deletion method. This method deletes vectors like vectors 51 to S3 shown in the figure, where one end point is in contact with vector a and whose length is shorter than a predetermined length and is considered to be noise, and converts it into a single vector a'. This is the way to do it.

As a result, the number of vectors can generally be reduced more than each of the linearization methods described above.

In this short vector deletion method, the length W of vectors considered as noise is a parameter, and as the value of the parameter W increases, the graphic accuracy decreases, but the amount of reduction in the number of vectors increases.

FIG. 2(d) shows replacement method (1). In this method, as shown in the figure, a set of vectors forming a character is replaced with a symbol Q indicating that the character is a character. However, this method is limited to cases where there is an instruction in the vectorization parameters, ``Character recognition is not performed and character candidates are replaced with symbols.'' This replacement method (1) It disappears, but
The number of vectors can be reduced to -a more than the linearization method and short vector deletion method described above.

FIG. 2(e) shows replacement method (2). As shown in the figure, this method converts a set of vectors with the fill attribute into a vector a indicating the outline of the fill area.
This method leaves only -C as a symbol and deletes the vector inside it. Similar to replacement method (1), this method is executed only when there is an instruction to "delete vectors inside the filled area" in the parameters.

When a triangular figure with a fill attribute is subjected to line thinning processing, it becomes a complex vector line diagram containing many vectors, as illustrated in the upper part of the figure (e). Therefore, by converting such a vector diagram into vector data of only its outline, it is generally possible to reduce the number of vectors more than in each of the above-mentioned methods.

The order in which the vector number reduction processes illustrated in Figure 2 are executed depends on the difference between the number of vectors obtained by the vectorization process and the set upper limit of the number of vectors, the required graphic accuracy, and the reading It is determined in advance through experiments, calculations, etc., depending on the amount of noise in the target drawing, the importance of character information, etc.

For example, if the difference between the number of vectors obtained by the vectorization process and the set upper limit of the number of vectors is small, linearization methods (1) and (2) are sequentially executed. This makes it possible to reduce the number of vectors while maintaining good accuracy.

Also, if the difference between the number of vectors obtained by vectorization processing and the upper limit of the set number of vectors is large,
The short vector deletion method and replacement methods (1) and (2) are executed first, and when the number of vectors approaches the upper limit, the linearization methods (1) and (2) are executed.

(5) Every time the vector number reduction process is completed, the process returns to step S3 to compare the number of vectors after the vector number reduction process with the upper limit number, and if the former is greater than the latter upper limit number, step S4 described above The vector number reduction process is executed again.

Thereafter, in the same manner, the vector number reduction processes in steps S and S4 are repeatedly executed until the number of vectors obtained by the vector number reduction process becomes equal to or less than the specified upper limit number.

When the number of vectors after the vector number reduction process becomes less than or equal to the specified upper limit, the vector number reduction process is terminated.
The data is output to a host computer (not shown), etc.

As described above, the upper limit of the final number of vectors obtained by the vector number reduction process is set, and the vector number reduction process is repeatedly executed until the number of vectors obtained by the vector number reduction process becomes less than or equal to the specified upper limit. As a result, the process of reducing the number of vectors to the specified upper limit is automatically executed, and the process of reducing the number of vectors can be speeded up.

In addition, there is no need for the operator to set or modify each parameter for vector number reduction processing.
It is possible to reduce the burden on the operator and improve work efficiency.

〔Example〕

An embodiment of the present invention will be described with reference to FIGS. 2 and 3. FIG. 3 shows an automatic drawing input system used to implement an embodiment of the present invention.

(A) Structure of automatic drawing input system In Fig. 3, 12 is an image scanner, and the figure (
(including characters).

13 is a preprocessing unit, which binarizes image data;
Value conversion processing, image quality selection processing to remove various noises from binarized image data, processing to extract features of the image data, and line thinning processing to narrow the line width of each figure whose image quality has been improved to obtain a center line. .

A vectorization processing section 14 performs vectorization processing to convert the image data of each figure thinned by the preprocessing section 13 into vector data.

Reference numeral 15 denotes a separation processing section, which performs a process of separating vector data into vector data of characters, symbols, filled figures, and line segments.

A post-processing unit 16 performs a vector number reduction process to reduce the number of hectors to a specified upper limit number or less.

In the post-processing section 16, 161 is a first linearization processing section, which performs vector number reduction processing using the linearization method (1). Reference numeral 162 denotes a second linearization processing unit, which performs vector number reduction processing using the linearization method (2).

A short vector deletion processing unit 163 performs vector number reduction processing using a short hector deletion method. A first replacement processing unit 164 performs vector number reduction processing using replacement method (1). Reference numeral 165 denotes a second replacement processing unit, which performs vector number reduction processing using replacement method (2).

17 is a main processor, which controls each process performed by the pre-processing section 13, the vectorization processing section 14, the separation processing section 15, and the post-processing section 16.

Reference numeral 18 denotes a memory unit, which includes a control program area 181 in which control programs for controlling each process performed by the preprocessing unit 13, vectorization processing unit 14, separation processing unit 15, postprocessing unit 16, and main processor 17 are stored; A data area 182 that stores various data such as data, a parameter area 183 that stores various parameters for vectorization processing and post-processing, and a counter area 1 that stores the number of vectors obtained by vector processing and vector number reduction processing.
84, each work area 185 used for processing each part is provided.

(B) Operation of the Embodiment The operation shown in FIG. 3 will be explained in the order of operation with reference to FIG. 2. The following operations performed in each section are executed by the control program stored in the control program area 181 of the memory section 1 under the control of the main processor 17, but this will not be mentioned unless it is particularly necessary. .

(1) A host computer (not shown) specifies various parameters necessary for vectorization processing and vector number reduction processing, and each parameter is stored in the parameter area 183 of the memory unit 18. Parameters for vector number reduction processing include linearization methods (1) and (2), short hector deletion method, and replacement methods (1) and (2).
) are specified.

Further, the upper limit number of vectors is stored in the counter area 184 of the memory unit 18.

(2) The image scanner 12 reads figures (including characters) on the drawing 11 and stores the image data in the data area 182 of the memory unit 18.

(3) The preprocessing unit 13 extracts image data from the data area 182, performs a binarization process to binarize the image data, an image quality selection process to remove various noises from the binarized image data, and features of the image data. Then, a line thinning process is performed to narrow the line width of each figure whose image quality has been improved to obtain a center line.

This thinned image data is stored in the memory section 18.
It is stored in the data area 182 of .

(4) The vectorization processing unit 14 performs vectorization processing to convert the image data of each figure thinned by the preprocessing unit 13 into vector data, and stores the obtained vector data in the data area 182 of the memory unit 18. Store.

(5) The separation processing unit 15 extracts vector data from the data area 182 and performs a process of separating the vector data into vector data of characters, symbols, line segments, and filled figures using various known methods. Characters are separated by features such as isolated vector data sequences consisting of a number of short line segments within a small frame area, for example.

The symbols are separated, for example, by matching with templates for each symbol.

Filled shapes are separated by the characteristic of isolated vector data sequences consisting of a number of short line segments, for example larger than a character frame and within an irregular area.

Through this separation process, the number of characters, symbols, and filled figures, as well as the number of vectors of all line segments, are detected.

(6) The main processor 17 performs control based on the number of characters, symbols, and filled figures separated by the separation processing unit 15, the number of vectors for all line segments, and the number of vectors set in the counter area 184 of the memory unit 18. Select the vector number reduction processing program in the program area 181.

Next, the main processor 17 executes the first linearization processing section 161, second linearization processing section 162, similar vector reduction processing section 163, and first replacement processing section 164 of the post-processing section 16 according to the vector number reduction processing program. and instructs one of the second replacement processing units 165 to execute the vector number reduction process using the method explained in FIG.

(7) Each time the vector number reduction process is completed, the main processor 17 compares the number of vectors after the vector number reduction process with the upper limit number in the counter area 184, and if the former is larger than the latter upper limit, , the next vector number reduction process is executed again according to the vector number reduction process program described above.

Thereafter, in the same manner, the above-described vector number reduction process is repeatedly executed until the number of vectors obtained by the vector number reduction process becomes equal to or less than the specified upper limit number.

When the number of vectors after the vector number reduction process becomes less than or equal to the specified upper limit, the vector number reduction process is terminated.
The data is output to a host computer (not shown), etc.

Although one embodiment of the present invention has been described above, the present invention is not limited to this embodiment, and various modifications can be made in accordance with the gist of the invention.

For example, as the vector number reduction processing method, various known vector number reduction processing methods other than the method shown in FIG. 2 can be used.

〔effect〕

As explained above, according to the present invention, the following effects can be obtained.

(1) Set the upper limit of the final number of vectors obtained by the vector number reduction process, and repeat the vector number reduction process until the number of vectors obtained by the vector number reduction process becomes less than or equal to the specified upper limit. Therefore, the process of reducing the number of vectors is automatically executed to the specified upper limit, and the process of reducing the number of vectors can be speeded up.

(2) Since the operator no longer needs to set or modify each parameter for vector number reduction processing, the burden on the operator can be reduced and work efficiency can be improved.

[Brief explanation of drawings]

FIG. 1 is a diagram of the principle of the present invention. FIG. 2 is an explanatory diagram of the vector number reduction processing method of the present invention. FIG. 3 is an explanatory diagram of a drawing input system used to implement an embodiment of the present invention. The figure is an explanatory diagram of a conventional drawing input system, and FIG. 5 is an explanatory diagram of a conventional vector number reduction processing method. In FIG. 3, 11... drawing, 12... image scanner, 13...
... Preprocessing unit, 14... Vectorization processing unit, 15...
- Separation processing unit, 16... Post-processing unit, 17... Main processor, 18... Memory unit.

Claims (2)

[Claims] (1) In a vector number reduction processing method that reduces the number of vectors obtained by vectorization processing that converts image data read from a drawing into vector data, (a) Vector number reduction is used as a parameter when vectorizing image data. (b) Execute vectorization processing to convert image data into vector data; (c) Set the upper limit of the number of vectors that will be finally output by the processing; (d) If the number of vectors is greater than the specified upper limit, perform a predetermined vector number reduction process to reduce the number of vectors until the number of vectors becomes equal to or less than the upper limit. A vector number reduction processing method characterized by executing the following. (2) Vector number reduction processing methods include a linearization method that converts multiple straight lines into a single straight line, a short vector deletion method that deletes short vectors that are seen as noise, and a replacement that replaces characters or filled figures with predetermined symbols. 2. The vector number reduction processing method according to claim 1, wherein the vector number reduction processing method is any one of the methods or a combination thereof.