JP2021168007A - Graphic data generation method, graphic data generation apparatus, and program - Google Patents

Abstract

To provide a graphic data generation method, a graphic data generation apparatus, and a program that can convert a system flow diagram to graphic data accurately and at high speed without a risk of increasing work man-hours of engineers even if the generated graphic data become complicated.SOLUTION: A graphic generation method of an embodiment includes an extraction step, a recognition step, a detection step, and a conversion step. In the extraction step, a written matter is extracted from a system flow diagram. In the recognition step, an object is recognized from the written matter using a neural network. In the detection step, characters are detected from the written matter using optical character recognition, and a line is detected using morphology processing. In the conversion step, the object, the characters and the line are converted so as to be output as graphic data.SELECTED DRAWING: Figure 1

Description

Embodiments of the present invention relate to graphic data generation methods, graphic data generation devices and programs.

Graphic data of monitoring and control devices in buildings, water and sewage bureaus, dams, etc. are created by an engineer visually recognizing the system flow diagram that is the original draft of the graphic and using a dedicated tool. This method has a problem that the work man-hours of the engineer increase according to the complexity of the graphic data to be created.

Japanese Unexamined Patent Publication No. 2017-142637

The problem to be solved by the present invention is that there is little risk that the work manpower of the engineer will increase even if the created graphic data becomes complicated, and the graphic data capable of converting the system flow diagram into graphic data accurately and at high speed. The purpose is to provide a generation method, a graphic data generation device, and a program.

In order to achieve the above object, the graphic generation method of the embodiment includes an extraction step, a recognition step, a detection step, and a conversion step. In the extraction step, the items to be described are extracted from the system flow diagram. In the recognition step, an object is recognized from the items described by using a neural network. In the detection step, characters are detected from the items described by using optical character recognition, and lines are detected by using morphology processing. In the conversion step, objects, characters and lines are converted so that they can be output as graphic data.

The figure which shows the functional structure of the graphic data generation apparatus which concerns on embodiment. The figure which shows an example of the system flow diagram which concerns on embodiment. The figure which shows an example of the graphic which concerns on embodiment. The flowchart which shows the operation of the graphic data generation apparatus which concerns on embodiment. The figure which shows a part of the state which extracted the description item from the system flow diagram which concerns on embodiment. The figure which shows an example of the object which the graphic data generation apparatus which concerns on embodiment should convert into secondary data. The figure which shows the state which specified the object to be detected by the graphic data generation apparatus which concerns on embodiment. The figure which showed the method of determining the line to be detected by the graphic data generation apparatus which concerns on embodiment. The figure which shows the state which specified the line which should be detected by the graphic data generation apparatus which concerns on embodiment. The figure which shows the state which specified the character which should be detected by the graphic data generation apparatus which concerns on embodiment.

Hereinafter, embodiments for carrying out the invention will be described. The graphic data generation method and the graphic data generation device 1 according to the embodiment will be described with reference to FIGS. 1 to 9.

FIG. 1 is a diagram showing a functional configuration of the graphic data generation device 1 according to the embodiment.

FIG. 2 is a diagram showing an example of a system flow diagram according to the embodiment.

FIG. 3 is a diagram showing an example of graphic 3 according to the embodiment.

The graphic data generation device 1 automatically generates data for generating the graphic 3 based on a predetermined rule based on the system flow diagram shown in FIG. 2, which is the original draft of the graphic 3 to be displayed on the monitoring control device 2. do.

The monitoring and control device 2 is a device for smoothly performing monitoring and management work of buildings, water and sewage bureaus, dams, and the like. The monitoring control device 2 has a graphic display function capable of systematically grasping the state of equipment in the system to be monitored. The graphic display function associates a signal with equipment information with a systematic diagram that imitates the equipment of the system to be monitored, and changes the equipment imitating the diagram according to the signal to change the state of the equipment. It is a function to display to the administrator in a visually easy-to-understand manner. The figure displayed by using the graphic display function is called graphic 3.

For example, as shown in FIG. 3, in the graphic 3, each symbol is associated with a signal from a sensor attached to the equipment, and the display is changed according to the signal. As a result, the equipment manager can systematically grasp the status of the equipment. The symbol referred to here is a graphic, a character, a numerical value, or the like displayed on the graphic 3. The data for generating the graphic 3 is referred to as graphic data or secondary data as required. The graphic data needs to be in a format (for example, XML format) that can be read by the monitoring control device 2.

Graphic data is created according to the customer's request. Generally, an engineer visually recognizes a system flow diagram that is a draft of graphic 3 and arranges a symbol prepared in advance in a dedicated drawing tool to create graphic data that meets a customer's request. It is assumed that the system flow diagram is generated by a work different from the graphic creation.

The graphic data generation device 1 according to the present embodiment includes an extraction unit 11, a recognition unit 12, a detection unit 13, and a conversion unit 14.

The extraction unit 11 extracts the items to be described from the system flow diagram of the system to be monitored. The system flow diagram is read by, for example, the reading unit 21. The reading unit 21 is, for example, a design support tool such as a CAD terminal device or a scanner device. The reading unit 21 reads the system flow diagram and outputs the read data to the extraction unit 11 of the graphic data generation device 1. The system flow diagram may be input in any way as long as it can be input to the graphic data generation device 1 as data. Hereinafter, the data of the system flow diagram input to the graphic data generation device 1 will be referred to as primary data as needed. The extraction unit 11 extracts all the items to be described including lines, characters, and symbols on the drawing of the system flow diagram from the input data of the system flow diagram.

The recognition unit 12 recognizes the object 4 from the description items extracted by the extraction unit 11 by using the neural network 22 in which the object 4 appearing in the graphic 3 of the monitoring control device 2 is learned in advance. The object 4 is a specific symbol in the symbol appearing in the graphic 3 of the monitoring control device 2. The neural network 22 is a convolutional neural network model (CNN model) in which the object 4 is learned in advance.

The detection unit 13 detects a character used for the graphic 3 by optical character recognition (OCR) from the description items extracted by the extraction unit 11. Further, the detection unit 13 detects a line used for the graphic 3 by morphology processing from the description items extracted by the extraction unit 11.

The conversion unit 14 converts the characters and lines used for the object 4 recognized by the recognition unit 12 and the graphic 3 detected by the detection unit 13 so that they can be output as the graphic 3.

The operation of the graphic data generation device 1 is shown below.

FIG. 4 is a flowchart showing the operation of the graphic data generation device 1 according to the embodiment. The conversion from the primary data to the secondary data by the graphic data generation device 1 is performed in the following steps.

In step S101, the data of the system flow diagram of the system to be monitored is input to the graphic data generation device 1.

In step S102, the description items including lines, characters, and symbols are extracted from the data of the system flow diagram.

FIG. 5 is a diagram showing a part of a state in which the items to be described are extracted from the system flow diagram according to the embodiment.

In step S103, the object 4 is recognized from the items described in the system flow diagram by using the neural network 22 in which the object 4 appearing in the graphic 3 of the monitoring control device 2 is learned in advance. In step S103, a convolutional neural network model (CNN model) that has learned a specific symbol (object) that is the basis of the symbol is used, and the object 4 to be recognized is specified from the description items extracted in step S102. The object 4 to be recognized is an object 4 to be converted into secondary data by the neural network 22.

FIG. 6 is a diagram showing an example of an object 4 to be converted into secondary data by the graphic data generation device 1 according to the embodiment.

The neural network 22 is learned so that when the items described in the primary data column of FIG. 6 are input, the object 4 described in the secondary data column is output. The items and objects shown in FIG. 6 are examples, and all the symbols (items to be described) used in the system flow diagram are learned.

The neural network 22 can recognize a plurality of objects 4 of the same type at the same time. As shown in the second stage of FIG. 6, for example, seven Circuit Breakers are recognized at the same time. Further, the neural network 22 can recognize a plurality of types of objects 4 at the same time. As shown in the sixth row of FIG. 6, for example, Motor, Road Breaker, and Single Point are recognized at the same time.

FIG. 7 is a diagram showing a state in which the graphic data generation device 1 according to the embodiment has specified an object 4 to be detected.

In step S104, the line used for the graphic 3 of the monitoring control device 2 is detected. Here, among the items not recognized in step S103, a line to be detected as secondary data is selected based on a predetermined rule. The line to be detected as secondary data refers to the line that "indicates the flow of flow" shown in the system flow diagram. The line may be detected by using an image conversion method (morphology processing) such as a morphology calculation.

An example of a given rule is to detect only straight lines that are orthogonal to each other. Another example of the predetermined rule is to detect only a line that is orthogonal to each other and connects the object 4 and the object 4 in three or less sections. Here, the "section" is a unit representing the number of line segments in a line composed of one line segment connecting objects or a line in which a plurality of line segments are sequentially connected. That is, in "one section", the number of line segments connecting objects is one. In the case of "two compartments", the number of line segments in a line in which a plurality of line segments connecting objects are sequentially connected is two. Similarly, in the "three compartments", the number of line segments in a line in which a plurality of line segments connecting objects are sequentially connected is three. “Three compartments or less” refers to any of the above “one compartment”, “two compartments”, and “three compartments”.

Predetermined rules may be preset in the graphic data generation device 1.

In the present embodiment, the case where the rule "is a straight line orthogonal to each other and detects only the line connecting the object 4 and the object 4 in three sections or less" will be described.

FIG. 8 is a diagram showing a method of determining a line to be detected by the graphic data generation device 1 according to the embodiment.

FIG. 8A shows a state in which only straight lines orthogonal to each other are detected from the items described in the system flow diagram extracted in step S102. However, in FIG. 8A, the undetectable straight line 5 used for the ruled lines of the table and the decoration of the figure in the system flow diagram is also extracted. Therefore, as shown in FIG. 8B, only the straight line 6 connected to the object 4 recognized in step S103 is detected from the straight line extracted in FIG. 8A. Next, as shown in FIG. 8 (c), a straight line 7 that is in contact with the straight line 6 detected in FIG. 8 (b) at two or more points is extracted from FIG. 8 (a) as a line to be detected. By this method, it is possible to "detect only the lines that are orthogonal to each other and connect the object 4 and the object 4 in three or less sections".

FIG. 9 is a diagram showing a state in which the graphic data generation device 1 according to the embodiment has specified a line to be detected.

In step S105, the character used for the graphic 3 of the monitoring control device 2 is detected by optical character recognition.

In step S106, the object 4 recognized in step S103, the line detected in step S104, and the character detected in step S105 are converted into graphic 3 so that they can be output. Convert to a specific format that can be output as graphic 3.

In this embodiment, by using the graphic data generation method and the graphic data generation device 1 described above, there is little risk that the work man-hours of the engineer will increase even if the created graphic data becomes complicated, and the system flow diagram can be obtained. It is possible to convert to graphic data accurately and at high speed.

When a system using the technique described in the present embodiment (hereinafter referred to as an automatic conversion system) is used, the secondary data automatically converted by the automatic conversion system can be obtained by inputting the primary data into the automatic conversion system. .. By performing the conversion work that was conventionally performed by engineers in the system, there is little risk that the amount of work will increase even if the secondary data to be created becomes large and complicated.

Although some embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other embodiments, and various omissions, replacements, and changes can be made without departing from the gist of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are also included in the scope of the invention described in the claims and the equivalent scope thereof.

1 ... Graphic data generation device 2 ... Monitoring control device 3 ... Graphic 4 ... Objects 5, 6, 7 ... Straight line 11 ... Extraction unit 12 ... Recognition unit 13 ... Detection unit 14 ... Conversion unit 21 ... Reading unit 22 ... Neural network

Claims (7)

Extraction steps to extract the items to be described from the system flow diagram, and
A recognition step that recognizes an object from the above items using a neural network,
From the above description, a detection step of detecting a character using optical character recognition and detecting a line using morphology processing, and a detection step.
A conversion step for converting the object, the character, and the line so that they can be output as graphic data, and
Graphic data generation method having. In the detection step, the lines to be detected are straight lines orthogonal to each other.
The graphic data generation method according to claim 1. In the detection step, the line to be detected is a line connecting the objects in three or less sections.
The graphic data generation method according to claim 2. An extraction unit that extracts items to be described from the system flow diagram,
A recognition unit that recognizes objects from the above items using a neural network,
From the above items, a detection unit that detects characters using optical character recognition and detects lines using morphology processing, and
A conversion unit that converts the object, the character, and the line so that they can be output as graphic data.
A graphic data generator equipped with. The lines detected by the detection unit are straight lines orthogonal to each other.
The graphic data generation device according to claim 4. The line detected by the detection unit is a line connecting the objects in three or less sections.
The graphic data generation device according to claim 5. On the computer
Extraction steps to extract the items to be described from the system flow diagram, and
A recognition step that recognizes an object from the above items using a neural network,
From the above description, a detection step of detecting a character using optical character recognition and detecting a line using morphology processing, and a detection step.
A conversion step for converting the object, the character, and the line so that they can be output as graphic data, and
A program to execute.

Images