Patent drawing creation apparatus, patent drawing creation method, and patent drawing creation program

Abstract

This invention provides a patented drawing creation device that is more efficient than manually creating drawings and produces higher quality drawings than those created using a single generation method, by applying the optimal generation method to various drawing types. [Solution] The patent drawing creation apparatus of the present invention comprises: a receiving unit that receives information about a drawing to be generated; a first image creation unit that creates a patent drawing using AI image generation technology based on the information received by the receiving unit; a second image creation unit that creates a code for drawing generation based on the information received by the receiving unit and creates a patent drawing from the code; a determination unit that determines whether to use the drawing creation method of the first image creation unit or the second image creation unit based on the information received by the receiving unit; and an output unit that outputs the patent drawing created by the first image creation unit or the second image creation unit.

Description

Technical Field

[0001] The present invention relates to a patent drawing creation device, a patent drawing creation method, and a patent drawing creation program.

Background Art

[0002] In recent years, drawing creation in patent applications has become an important element for visually explaining the content of an invention. There are various types of drawings in patent drawings, such as usage diagrams, UI diagrams, flowcharts, block diagrams, etc., and appropriately creating these has become an important element that affects the quality of patent applications.

[0003] Conventionally, these patent drawings have generally been created manually. However, manual drawing creation has the problems of requiring a huge amount of time and having variations in quality depending on the skills of the creator. For this reason, technologies for improving the efficiency and stabilizing the quality of drawing creation are demanded.

[0004] In recent years, with the development of artificial intelligence technology, drawing creation methods using AI image generation technology and code generation technology have been proposed. While AI image generation technology is excellent in realistic expression and drawing of complex scenes, code generation technology is suitable for creating drawings with logical structures such as flowcharts and block diagrams.

[0005] Patent Document 1 discloses an information processing device that creates a drawing showing the configuration from the claims using a computer. In this technology, the claims are input, and code for drawing creation is obtained by processing with a large language model, and a block diagram or flowchart is drawn according to the code.

Prior Art Documents

Patent Documents

[0006]

Patent Document 1

Summary of the Invention

[0007] However, Patent Document 1 employs only a code-based drawing generation method, which has the problem of not being able to adequately handle drawing types that require realistic representation, such as the diagram used. Furthermore, Patent Document 1 is a technology specialized in drawing block diagrams or flowcharts, and it was difficult to provide optimal quality for all diverse drawing types with only a single generation method.

[0008] Therefore, the present invention aims to provide a patentable drawing creation device that is more efficient than manually creating drawings and improves quality compared to drawing created with a single creation method, by applying an optimal generation method to various types of drawings. [Means for solving the problem]

[0009] To solve the above problems, the patented drawing creation apparatus of the present invention includes a receiving unit that receives information about the drawing to be generated, A first image creation unit creates patent drawings using AI image generation technology based on the information received at the aforementioned reception unit, A second image creation unit creates a code for generating drawings based on the information received by the reception unit and creates patent drawings from the said code. A determination unit determines, based on the information received by the reception unit, whether to use the drawing creation method of the first image creation unit or the second image creation unit, An output unit that outputs patent drawings created by the first image creation unit or the second image creation unit 、 Based on the information received at the aforementioned reception unit, a drawing type acquisition unit acquires the type of drawing, Equipped with 、 The determination unit determines, according to the type of drawing, whether to use the drawing creation method of the first image creation unit or the second image creation unit. It is characterized by the following: [Effects of the Invention]

[0010] According to the present invention, by applying the optimal generation method to various types of drawings, it is possible to provide a patentable drawing creation device that is more efficient than manually creating drawings and improves quality compared to creating drawings with a single generation method. [Brief explanation of the drawing]

[0011] [Figure 1] This is a block diagram showing an example of the overall configuration of the patent drawing creation device according to this embodiment. [Figure 2] This flowchart shows an example of the processing flow in the patent drawing creation device according to this embodiment. [Figure 3] This figure shows an example of an input user interface according to this embodiment. [Figure 4] This figure shows an example of a drawing created by the first image creation unit according to this embodiment. [Figure 5] This figure shows an example of a drawing created by the second image creation unit according to this embodiment. [Figure 6] This figure shows an example of another drawing created by the second image creation unit according to this embodiment. [Figure 7] This figure shows an example of code created by the second image creation unit according to this embodiment. [Figure 8] This figure shows another example of the input user interface according to this embodiment. [Modes for carrying out the invention]

[0012] The following description of this embodiment will be made with reference to the drawings. It should be noted that the present invention is not limited to the embodiments shown below, and can be modified, added, altered, or deleted to the extent that a person skilled in the art can conceive of it. Any embodiment that achieves the function and effect of the present invention is included within the scope of the present invention.

[0013] The patent drawing creation device according to this embodiment includes a reception unit that receives information regarding the drawing to be generated, a first image creation unit that creates a patent drawing by an AI image generation technique based on the information received by the reception unit, a second image creation unit that creates code for drawing generation based on the information received by the reception unit and creates a patent drawing from the code, a determination unit that determines whether to use the drawing creation method of either the first image creation unit or the second image creation unit based on the information received by the reception unit, and an output unit that outputs the patent drawing created by the first image creation unit or the second image creation unit. The patent drawing creation device of this embodiment may be configured as a system.

[0014] Further, it is preferable to include a drawing type acquisition unit that acquires the type of the drawing based on the information received by the reception unit, and the determination unit determines whether to use the drawing creation method of either the first image creation unit or the second image creation unit according to the type of the drawing.

[0015] Further, there are a plurality of drawing creation methods in the first image creation unit, and it is preferable that the determination unit determines whether to use any one of the plurality of drawing creation methods in the first image creation unit or the drawing creation method in the second image creation unit based on the information received by the reception unit.

[0016] Further, there are a plurality of drawing creation methods in the second image creation unit, and it is preferable that the determination unit determines whether to use the drawing creation method in the first image creation unit or any one of the plurality of drawing creation methods in the second image creation unit based on the information received by the reception unit.

[0017] Further, it is preferable that the drawing type acquisition unit acquires the type of the drawing by input or selection from the user, or estimates the type of the drawing by an AI generation model based on the information received by the reception unit.

[0018] Furthermore, the patent drawing creation method according to this embodiment is a computer-based patent drawing creation method, and is characterized by comprising: a reception step of receiving information about a drawing to be generated; a first image creation step of creating a patent drawing using AI image generation technology based on the information received in the reception step; a second image creation step of creating a code for drawing generation based on the information received in the reception step and creating a patent drawing from the code; a determination step of determining whether to use the drawing creation method of the first image creation step or the second image creation step based on the information received in the reception step; and an output step of outputting the patent drawing created in the first image creation step or the second image creation step.

[0019] Furthermore, the patent drawing creation program according to this embodiment is characterized by causing a computer to execute: an acceptance process for receiving information about a drawing to be generated; a first image creation process for creating a patent drawing using AI image generation technology based on the information received in the acceptance process; a second image creation process for creating a code for drawing generation and creating a patent drawing from the code based on the information received in the acceptance process; a determination process for determining whether to use the drawing creation method of the first image creation process or the second image creation process based on the information received in the acceptance process; and an output process for outputting the patent drawing created by the first image creation process or the second image creation process.

[0020] Figure 1 is a block diagram showing an example of the overall configuration of a patent drawing creation device according to this embodiment. As shown in Figure 1, the patent drawing creation device 1 includes a reception unit 11 that receives information about the drawing to be generated, a first image creation unit 13 that creates a patent drawing using AI image generation technology based on the information received by the reception unit 11, a second image creation unit 14 that creates a code for drawing generation based on the information received by the reception unit 11 and creates a patent drawing from the code, a determination unit 12 that determines whether to use the drawing creation method of either the first image creation unit 13 or the second image creation unit 14 based on the information received by the reception unit 11, and an output unit 15 that outputs the patent drawing created by the first image creation unit 13 or the second image creation unit 14. The patent drawing creation device 1 receives information necessary for drawing generation from a user 20 via a computer terminal, the determination unit 12 selects the optimal creation method according to the characteristics and content of the drawing, and the selected first image creation unit 13 or second image creation unit 14 automatically generates the patent drawing. The patent drawing creation device 1 may include a drawing type acquisition unit 16 that acquires the type of drawing based on information received by the reception unit 11. The drawing type acquisition unit 16 may, for example, acquire the type of drawing based on input or selection from the user 20, or estimate the type of drawing using an AI generation model based on information received by the reception unit 11.

[0021] The reception unit 11 receives information from the user 20 regarding the drawing to be generated. As shown in Figure 1, the reception unit 11 functions as an input interface for the patent drawing creation device 1 and acquires various information entered by the user 20 through a computer terminal. The information received by the reception unit 11 may include the content of the drawing to be created, the type of drawing, elements to be included in the drawing, and technical features. The reception unit 11 may support multiple input methods, such as text input, voice input, file upload (e.g., image data), or selection by the user 20. The information received by the reception unit 11 is sent to the determination unit 12 for use in determining the drawing creation method, or sent to the drawing type acquisition unit 16 for use as material for identifying the type of drawing. Furthermore, the information received by the reception unit 11 is also used as basic data for drawing creation in the first image creation unit 13 or the second image creation unit 14.

[0022] Figure 3 shows an example of a drawing information input screen provided by the reception unit 11. In Figure 3, for example, the title "Drawing Information Input Screen" is displayed at the top of the screen, and below it is an input instruction message that reads "Please enter information about the drawing." In the center of the screen, there is a text input field for the user 20 to enter information about the drawing to be generated, and this field displays a specific example description: "A drawing of an invention in which a drone holds an umbrella." The information entered in this text input field is received by the reception unit 11 as information 31 about the drawing to be generated. The user 20 can freely describe the content of the drawing they wish to create, the type of drawing, the components to be included in the drawing, technical features, etc. An "Execute" button is located in the lower right corner of the screen, and when the user 20 presses this button after completing the input, the entered information is sent to the reception unit 11.

[0023] The determination unit 12 determines, based on the information received by the reception unit 11, whether to use the drawing creation method of either the first image creation unit 13 or the second image creation unit 14. The determination of the drawing creation method may involve either using a drawing type or not using a drawing type. If drawing types are not used, the determination unit 12 makes a determination using AI (which may be called, for example, a generative AI, a trained model, a generative model, a large-scale language model, etc.). Specifically, the determination unit 12 inputs the received information into the AI ​​and, based on the analysis results by the AI, decides which drawing creation method of the first image creation unit 13 or the second image creation unit 14 to use. For example, the determination unit 12 inputs a prompt to the AI ​​saying, "Based on the information received, please determine whether structural drawings are required or drawings with a high degree of flexibility are required." In this case, the determination is made directly from the content of the information received, without using classifications of drawings (e.g., block diagrams, circuit diagrams, flowcharts, etc.).

[0024] The determination unit 12 may determine whether to use the drawing creation method of either the first image creation unit 13 or the second image creation unit 14, depending on the type of drawing acquired by the drawing type acquisition unit 16. Possible types of drawings include block diagrams, flowcharts, structural diagrams, and circuit diagrams, but are not limited to these. The determination unit 12 uses the drawing type information acquired from the drawing type acquisition unit 16 to select a drawing creation method suitable for each type. For example, in the case of a diagram of a machine in use, the determination unit 12 may select the first image creation unit 13 which uses AI image generation technology, and in the case of a block diagram showing a system configuration, it may select the second image creation unit 14 which creates the drawing from code.

[0025] The determination unit 12 may, when the first image creation unit 13 has multiple drawing creation methods, determine, based on the information received by the reception unit 11, whether to use one of the multiple drawing creation methods of the first image creation unit 13 or a drawing creation method of the second image creation unit 14. When the first image creation unit 13 has multiple AI image generation technologies, options include, for example, a method using a diffusion model, a method using a generative adversarial network, and a method using a variational autoencoder. Furthermore, even if the same method is used, if the learning method or parameters differ, it is conceivable to use them accordingly. Alternatively, publicly available AI generation services may be used. By using multiple drawing creation methods in this way, the optimal method can be selected according to the strengths and weaknesses of each generation method, thereby enabling the generation of higher quality drawings.

[0026] The determination unit 12 may, when the second image creation unit 14 has multiple drawing creation methods, determine, based on the information received by the reception unit 11, whether to use the drawing creation method of the first image creation unit 13 or one of the multiple drawing creation methods of the second image creation unit 14. When the second image creation unit 14 has multiple code-based drawing creation methods, possible options include, for example, a method for generating drawings in vector format, a method for generating drawings in raster format, a method for using a specific drawing description language, a method for drawing using SVG (Scalable Vector Graphics), and a method for drawing by combining HTML, CSS, and JavaScript. The first image creation unit 13 may be equipped with multiple AI image generation methods, and the second image creation unit 14 may be equipped with multiple code-based drawing creation methods. In this case, the determination unit 12 selects the drawing creation method to be used from among the multiple methods of the first image creation unit 13 and the multiple methods of the second image creation unit 14, based on the information received by the reception unit 11. By using multiple drawing creation methods in this way, it becomes possible to select the most suitable method according to the strengths and weaknesses of each generation method, thereby enabling the creation of higher-quality drawings. For example, if there is a performance difference between the AI ​​image generation technology and the code generation technology available to the user, and the AI ​​image generation technology is more accurate in creating usage diagrams than the code generation technology, then it is desirable to choose the AI ​​image generation technology for drawings that vary widely depending on the type of machine or device and the viewing angle, such as usage diagrams for machinery and equipment. On the other hand, in diagrams such as block diagrams, when code generation technology is advantageous, for example, when generating text information from code makes it easier to accurately reflect the text information, it is desirable to create the diagram using code generation technology. However, the drawing creation method to be selected is not limited to the above, and will be appropriately selected depending on the AI ​​model, coding technology, invention content, patent content, type of drawing, etc.

[0027] The first image creation unit 13 creates patent drawings using generation AI technology based on the information received by the reception unit 11. Specifically, the first image creation unit 13 uses the information received from the reception unit 11 to configure prompts for image generation and inputs them into a trained AI image generation model to generate patent drawings. Examples of prompts include natural language instructions such as "Please draw a diagram of a drone holding an umbrella." The first image creation unit 13 may use the information received by the reception unit 11 as prompts as is, or it may process the received information to convert it into more detailed prompts before inputting them into the image generation model. For example, based on the received summary of the invention and technical features, it may generate prompts that include information specifying the type of drawing (overall view, cross-sectional view, perspective view, usage view, etc.). Furthermore, the first image creation unit 13 may generate multiple different prompts to create multiple variations of drawings, allowing the user to select from among them. In addition, the first image creation unit 13 may receive feedback from the user on the generated images, modify the prompts, and perform an iterative process to generate the images again. The drawing generated by the first image creation unit 13 is provided to the user through the output unit 15.

[0028] The second image creation unit 14 creates a code for generating drawings based on the information received by the reception unit 11, and creates patent drawings from this code. As shown in Figure 2, the second image creation unit 14 operates selectively according to the determination result by the determination unit 12 and generates drawings using a different method than the first image creation unit 13. Specifically, the second image creation unit 14 analyzes the drawing information received by the reception unit 11 and generates a code for generating drawings that defines the components, arrangement, and connection relationships of the drawings. There are two methods for generating this code: one that generates the code from scratch, and another that uses pre-prepared code templates. In the template method, for example, a template code that defines the basic structure of a drawing is prepared in advance, and only the variable parts according to the content of the drawing are generated by the generation AI. For example, in a block diagram template, a basic code that defines the arrangement of blocks and the structure of connecting lines is prepared, and the entire code can be completed by generating only the variable elements such as text and symbols to be displayed within each block using the generation AI. By using this template method, the basic structure of the drawing can be maintained while flexibly changing only the necessary parts. One technique for generating code is to use a generation AI. For example, by inputting a prompt to the generation AI such as "Generate code to draw the user interface screen of an application for controlling a drone," it is possible to obtain code for generating diagrams. In the template method, for example, by inputting a prompt to the generation AI such as "Generate text to be displayed in each block in a block diagram showing the system configuration," it is possible to obtain text information to be inserted into the variable parts of the template code and complete the code by incorporating this into the template code. The generated code is written, for example, in a graphic drawing language or a programming language, and the execution of this code generates the patent drawing as an image. The code generation method by the second image creation unit 14 has the characteristic of increasing the accuracy of the textual information contained in the drawing. This is because, since the string can be directly specified by the code, textual information such as symbols, explanatory text, and numerical values ​​is clearly and accurately placed on the drawing. The generated patent drawing is provided to the user via the output unit 15.

[0029] The output unit 15 outputs the patent drawings created by the first image creation unit 13 or the second image creation unit 14. As shown in Figure 1, the output unit 15 is arranged as a component of the patent drawing creation device 1 and provides the drawing data generated by the first image creation unit 13 or the second image creation unit 14 in a format usable by the user 20. The output unit 15 provides the created patent drawings in output forms such as displaying them on a screen, printing them, or saving them as a file.

[0030] The drawing type acquisition unit 16 acquires the type of drawing to be generated based on the information received by the reception unit 11. The determination unit 12 decides whether to use the first image creation unit 13 or the second image creation unit 14, depending on the type of drawing acquired by the drawing type acquisition unit 16. Specifically, the types of drawings acquired by the drawing type acquisition unit 16 include system configuration diagrams, block diagrams, flowcharts, circuit diagrams, cross-sectional views of machinery, side views of machinery, top views of machinery, perspective views of parts, exploded views of parts, chemical structural formulas, graphs, time charts, network configuration diagrams, data structure diagrams, screen transition diagrams, sequence diagrams, class diagrams, layout diagrams, process diagrams, manufacturing process diagrams, assembly diagrams, operation explanation diagrams, state transition diagrams, relationship diagrams, conceptual diagrams, etc.

[0031] Furthermore, the drawing type acquisition unit 16 may acquire the drawing type through user input or selection, or it may estimate the drawing type using an AI generation model based on information received by the reception unit 11. In the user input acquisition method, the user directly specifies the type of drawing they wish to create through a selection interface such as a pull-down menu or radio buttons on the input screen provided by the reception unit 11. For example, if the user selects the desired drawing type from options such as "block diagram," "flowchart," or "usage diagram," the drawing type acquisition unit 16 acquires the selection information and provides it to the determination unit 12. On the other hand, in the AI ​​generation model estimation method, the drawing type acquisition unit 16 inputs the drawing information received by the reception unit 11 into the AI ​​generation model, analyzes the content of the information, and automatically determines the appropriate drawing type. For example, if information such as "shows the relationships between each component of the system" is received, the AI ​​generation model classifies it as a block diagram, and the drawing type acquisition unit 16 acquires the estimation result. Also, if information such as "shows the operating procedure of the device" is received, it is classified as a flowchart. In this estimation process, the drawing type acquisition unit 16 may input instructions to the AI ​​generation model as a prompt, such as "Determine the type of drawing to be created from the following information. Options include block diagrams, flowcharts, usage diagrams, structural diagrams, etc."

[0032] User 20 is a person who uses this device to create patent drawings, as shown in Figure 1. User 20 may be, but is not limited to, an inventor, a patent office employee, or a person in the intellectual property department of a company.

[0033] The information 31 regarding the drawing to be generated is information (such as text information) that indicates the content of the patent drawing that the user 20 wishes to create. In the text input field in Figure 3, the user 20 enters a sentence as the information 31 regarding the drawing to be generated, for example, "A diagram of the use of an invention in which a drone holds an umbrella." The information 31 regarding the drawing to be generated is received by the reception unit and used in the determination of the drawing creation method in the determination unit and in the patent drawing creation process in the first image creation unit and the second image creation unit. As shown in Figure 8, the information 31 regarding the drawing to be generated may also be entered in combination with an option to specify the type of drawing. In this case, the user 20 provides a description of the invention, such as "An invention in which a drone holds an umbrella," along with a specification of the type of drawing, such as a block diagram, flowchart, UI diagram, or usage diagram.

[0034] Patent drawings 32 are drawing data generated in this system and are used, for example, as drawings for patent applications, utility model registration applications, design registration applications, international patent applications, drawings for technical documents shared within the company, drawings for invention proposals, drawings for technical presentation materials, explanatory drawings for license negotiations, drawings for technology transfer materials, and drawings for shared materials with the product development department. Specific examples of patent drawings 32 include perspective views showing the structure of a machine or device, block diagrams showing system configurations, flowcharts showing processing procedures, front views showing user interface screens, circuit diagrams showing electrical circuits, structural diagrams showing chemical structures, process diagrams showing manufacturing processes, cross-sectional views showing cross-sections of devices, plan views showing component layouts, side views showing operating states, conceptual diagrams showing data structures, network diagrams showing network configurations, timing charts showing time-series changes, or explanatory diagrams showing physical phenomena. As shown in Figure 4, patent drawing 32 is an example created by the first image creation unit, generated as a perspective view showing a configuration in which an umbrella is attached to a drone, and is depicted in a line drawing style in which the drone body, propellers, umbrella, and support are clearly distinguishable. As shown in Figure 5, patent drawing 32 is an example created by the second image creation unit, generated as a user interface screen for a drone control system, and includes a drone status display area, control buttons, a flight control panel, and a warning notification area. As shown in Figure 6, patent drawing 32 is an example created by the second image creation unit, generated as a flowchart including the steps of image acquisition, image analysis, position information calculation, movement direction amount determination, and drive unit transmission. Figure 7 is an example of drawing generation code created by the second image creation unit to generate the flowchart in Figure 6. Patent drawing 32 is generated by either the first image creation unit using AI image generation technology, or the second image creation unit creating drawing generation code and generating the drawing from that code, and is output by the output unit.

[0035] Drawing type 33 is information indicating the type of drawing produced by the patent drawing creation device. Specifically, type 33 of drawings can include system configuration diagrams, block diagrams, flowcharts, circuit diagrams, cross-sectional views of machinery, side views of machinery, top views of machinery, front views of machinery, rear views of machinery, bottom views of machinery, perspective views of parts, exploded views of parts, enlarged views of parts, detailed views of parts, chemical structural formulas, molecular structural diagrams, reaction scheme diagrams, graphs, time charts, waveform diagrams, network configuration diagrams, communication sequence diagrams, data structure diagrams, data flow diagrams, screen transition diagrams, user interface diagrams, sequence diagrams, class diagrams, use case diagrams, activity diagrams, component diagrams, layout diagrams, process diagrams, manufacturing process diagrams, assembly diagrams, operation explanation diagrams, state transition diagrams, relationship diagrams, conceptual diagrams, schematic diagrams, principle explanation diagrams, comparison diagrams, example explanation diagrams, modified example explanation diagrams, cross-sectional structure diagrams, stacked structure diagrams, wiring diagrams, layout diagrams, optical system configuration diagrams, sensor placement diagrams, control block diagrams, signal processing flow diagrams, algorithm flow diagrams, decision flow diagrams, table configuration diagrams, memory map diagrams, etc. In this way, by using drawing type 33, a drawing creation method suitable for the content of the invention can be selected.

[0036] Figure 8 shows another example of the drawing information input screen. As shown in Figure 8, the drawing types 33 are presented in radio button format as options such as block diagram, flowchart, UI diagram, usage diagram, and "leave it to AI". The user selects the desired drawing type 33 along with inputting information 31 about the drawing to be generated, and the determination unit 12 selects an image creation unit corresponding to that type to create an appropriate patent drawing.

[0037] Figure 2 is a flowchart showing the processing steps of the patent drawing creation method executed by the patent drawing creation device 1. As shown in Figure 2, the patent drawing creation method according to this embodiment first receives information about the drawing to be generated in the reception step (S1). Next, in the determination step (S2), it is determined which of the first image creation step and the second image creation step to use based on the information received in the reception step (S1). Depending on the determination result of the determination step (S2), the process branches. In one branch, the first image creation step (S3) is executed, and the patent drawing is created using AI image generation technology. In the other branch, the second image creation step (S4) is executed, a code for drawing generation is created, and the patent drawing is created from the code. Once either the first image creation step (S3) or the second image creation step (S4) is completed, the created patent drawing is output in the output step (S5). After the output step (S5) is completed, the series of processes ends. In this way, the patent drawing creation method according to this embodiment can select an appropriate drawing creation method based on the received information and automatically create and output a patent drawing.

[0038] In the patent drawing creation apparatus according to this embodiment, multiple patent drawings may be created from information on a single drawing to be generated that is received by the reception unit. For example, if the text entered by the user contains multiple drawing elements, that is, if a description of a block diagram and a description of a diagram showing the state of use are included in a single input text, the first or second image creation unit will create multiple patent drawings corresponding to each drawing element. Alternatively, even if a single input text relates to a single drawing element, multiple drawing creation methods may be executed in parallel to create multiple patent drawings for the same drawing element. In this case, the determination unit selects the method to be applied from among the multiple available drawing creation methods based on the characteristics of the input drawing element. For example, by employing an algorithm in which the determination unit selects the two most optimal methods from 10 generation methods, efficient drawing creation according to the type and complexity of the drawing element becomes possible. Specifically, when information regarding the user interface screen is input, the determination unit selects a method suitable for the drawing element in question, either the AI ​​image generation technology method in the first image creation unit or the code generation method in the second image creation unit. Both selected methods are then executed, and the patent drawing created by the first image creation unit and the patent drawing created by the second image creation unit are output from the output unit. This allows the user to compare and examine multiple patent drawings and select the desired drawing.

[0039] Alternatively, the patent drawing creation device may include a modification receiving unit that receives modification instructions from the user for the created patent drawing, and a drawing modification unit that modifies the patent drawing based on the modification instructions. For drawings created by the first image creation unit, the drawing modification unit regenerates the drawing using AI image generation technology with the modification instructions as a prompt, and for drawings created by the second image creation unit, it modifies the drawing generation code to regenerate the drawing. This allows the user to adjust the created patent drawing to their desired form.

[0040] Alternatively, the patent drawing creation device may include a database that stores previously created patent drawings and their creation conditions, and a search unit that searches for past creation conditions similar to the information received by the reception unit. If the search unit finds similar past creation conditions, the determination unit preferentially selects the drawing creation method used for those past creation conditions. The database also stores evaluation results of the drawing quality for each drawing creation method, and the determination unit uses these evaluation results to select the drawing creation method.

[0041] In another form, the patent drawing creation device may include a quality evaluation unit that evaluates the quality of the created patent drawing. The quality evaluation unit quantifies and evaluates the clarity of lines in the patent drawing, the appropriateness of the placement of symbols, the overall balance of the drawing, and the degree of conformity to the Japan Patent Office's drawing rules. If the evaluation result by the quality evaluation unit falls below a predetermined threshold, the first or second image creation unit recreates the patent drawing using different parameters or a different method. This makes it possible to automatically generate patent drawings that meet certain quality standards.

[0042] Alternatively, the patent drawing creation device may include a consistency verification unit that verifies the consistency between multiple drawings. The consistency verification unit verifies whether the same reference numerals are assigned to the same components in multiple patent drawings included in the same patent application, and whether there are any inconsistencies in the representations between the drawings. If the consistency verification unit detects inconsistencies as a result of the verification, it identifies the inconsistent parts and notifies the user, or automatically makes corrections to unify the reference numerals.

[0043] In another form, the patent drawing creation device may include a code extraction unit that extracts codes and descriptions of codes to be assigned to the drawing from information received at a reception unit, and a code placement unit that places the extracted codes in appropriate positions on the patent drawing. The code extraction unit extracts the names of components and their corresponding code numbers from text information, and the code placement unit recognizes each component in the created patent drawing and places the corresponding codes along with leader lines. The code placement unit optimizes the placement positions so that codes do not overlap and leader lines do not intersect.

[0044] In another form, the patent drawing creation device may include a format conversion unit that converts the created patent drawings into different drawing formats. The format conversion unit performs, for example, conversion from an oblique view to a front view, a side view, and a plan view, or conversion from a physical drawing to a block drawing. For drawings created by the first image creation unit, the format conversion unit generates drawings from different viewpoints using AI image generation technology, and for drawings created by the second image creation unit, it generates drawings in different formats by changing the parameters of the code.

[0045] Generally, in the creation of patent drawings, adopting a single image generation method is preferred for its simplicity of design. Combining multiple generation methods leads to system complexity, resulting in increased development costs and maintenance burdens, so system developers usually avoid such configurations. However, the patent drawing creation device according to the present invention combines a first image creation unit using AI image generation technology with a second image creation unit that creates drawings from code, and further incorporates an automatic selection function by a determination unit. This enables not just the coexistence of methods, but dynamic switching of the optimal generation method according to the drawing type. This configuration yields synergistic effects that outweigh the disadvantages of system complexity caused by combining multiple methods.

[0046] According to the present invention, by applying the optimal generation method to various types of drawings, the creation time can be significantly reduced compared to manually creating drawings, thereby improving the efficiency of patent application work. Furthermore, compared to using only a single generation method to handle all types of drawings, it becomes possible to generate drawings that leverage the strengths of each generation method, thus improving the quality of the output drawings. For example, a second image creation unit using code generation can be used for drawings with regular structures such as flowcharts and block diagrams, while a first image creation unit using AI image generation can be used for drawings containing complex shapes such as external views of devices. This allows for generation that is suitable for each type of drawing.

[0047] Furthermore, in this invention, by using standard templates and libraries in drawing creation through code generation, variations in drawing quality are suppressed, enabling drawing generation with consistent quality standards. This promotes the standardization of patent drawings. In addition, since the judgment unit automatically determines the generation method based on the received information, users do not need specialized knowledge regarding the selection of drawing creation methods, and human resources can be effectively utilized by reducing manual work. Thus, the patent drawing creation device according to the present invention achieves both comprehensive support for drawing types and stable quality drawing output through the cooperation of a combination of multiple generation methods and an automatic judgment function. [Examples]

[0048] The present invention will be described in more detail below with reference to examples, but the present invention is not limited to these examples.

[0049] (Example 1) In this embodiment, with reference to Figures 3 and 4, a specific example of creating a patent drawing for an invention involving a drone holding an umbrella using the patent drawing creation device 1 will be described. First, the user 20 inputs information 31 regarding the drawing to be generated via the drawing information input screen shown in Figure 3. In this embodiment, the user 20 enters the string "Usage drawing of the invention involving a drone holding an umbrella" into the text input field. This input includes the object of the invention (a combination of a drone and an umbrella) and the desired type of drawing (usage drawing). When the user 20 presses the "Execute" button, the reception unit 11 receives the information and transmits it to the determination unit 12. Based on the received information, the determination unit 12 determines which drawing creation method, the first image creation unit 13 or the second image creation unit 14, to use. In this embodiment, since a drawing type called "usage drawing" and a drawing of a specific item called "invention involving a drone holding an umbrella" are required, the determination unit 12 selects the first image creation unit 13, which has AI image generation technology suitable for representing the appearance of a three-dimensional item. The first image creation unit 13 creates a patent drawing using AI image generation technology based on the input information. As shown in Figure 4, the created patent drawing 32 is a perspective view of a quadcopter-type drone equipped with four rotor propellers, with an open umbrella supported by a support strut at the top center. This patent drawing 32 is drawn in a line drawing style, and the structure of the drone body, propellers, landing gear, and umbrella are clearly distinguishable. The output unit 15 outputs the patent drawing 32 created by the first image creation unit 13 to the user 20. Thus, according to this embodiment, by simply inputting the concise information "a diagram of the invention in use with a drone holding an umbrella," the patent drawing creation device 1 can automatically select an appropriate drawing creation method and generate a patent drawing 32 that accurately represents the content of the invention. In this embodiment, the drawing type is obtained by the drawing type acquisition unit 16, but this is not the only example. Even if the drawing type acquisition unit 16 does not acquire the drawing type, the determination unit 12 can determine the drawing creation method based on the information received by the reception unit 11, thereby enabling the creation of patent drawings.

[0050] (Example 2) In this embodiment, the case of generating the UI diagram shown in Figure 5 will be described. User 20 inputs information 31 related to the diagram to be generated to the reception unit 11, which is "Operation screen of a drone control system". Based on the information received by the reception unit 11, the determination unit 12 determines that the diagram creation method of the second image creation unit 14 should be used. Based on the information received by the reception unit 11, the second image creation unit 14 creates a code for diagram generation and creates a patent diagram 32 from the code. As shown in Figure 5, the created patent diagram 32 has the title "Drone Control System", and the left panel includes a "Drone Status" section with display areas for altitude, GPS coordinates, wind speed, tilt angle, and battery level, and a "Control" section with "Open" and "Close" buttons, an angle adjustment slider, and a follow mode selection. The right panel includes a "Flight Control" section with joystick-type controls, an altitude display bar, an "Emergency Stop" button, and "Landing," "Takeoff," and "Hovering" buttons, and at the bottom is a "Warning / Notification" section with a display area for notification messages. The output unit 15 outputs the patent drawing 32 created by the second image creation unit 14. In this embodiment, the drawing type is obtained by the drawing type acquisition unit 16, but this is not the only example. Even if the drawing type acquisition unit 16 does not acquire the drawing type, the determination unit 12 determines the drawing creation method based on the information received by the reception unit 11, thereby enabling the creation of the patent drawing.

[0051] (Example 3) In this embodiment, the creation of a flowchart-style patent drawing by the second image creation unit 14 will be described with reference to Figures 6 and 7. Assume that user 20 inputs information 31 regarding the drawing to be generated, such as "a flowchart showing the processing flow of an image tracking system," into the input field. Based on this information, the determination unit 12 determines that the drawing creation method by the second image creation unit 14 is suitable for drawings that require a regular arrangement of shapes and arrow connections, such as a flowchart. Based on the received information, the second image creation unit 14 creates code for generating an SVG-style drawing as shown in Figure 7. This code includes drawing each processing step using rectangular elements, arranging text such as "Start," "Image Acquisition Step," "Image Processing Step," "Position Information Calculation Step," "Movement Direction Amount Determination Step," "Drive Unit Transmission Step," and "End" using text elements, and drawing arrows using path elements and polygon elements. By executing this SVG code, the second image creation unit 14 creates a flowchart-style patent drawing 32 as shown in Figure 6. The created flowchart has a structure in which each step is sequentially connected by arrows from start to finish, with the start and end represented by rounded rectangles and each processing step by a rectangle. The output unit 15 outputs the patent drawing 32 created in this way. In this embodiment, the drawing type is obtained by the drawing type acquisition unit 16, but it is not limited to this, and even if the drawing type acquisition unit 16 does not acquire the drawing type, the determination unit 12 can determine the drawing creation method based on the information received by the reception unit 11, thereby enabling the creation of the patent drawing.

[0052] (Example 4) In this embodiment, the selection of a drawing type using the user interface screen shown in Figure 8 will be described. As shown in Figure 8, the drawing information input screen is provided with a text input field for inputting information 31 about the drawing to be generated. The user 20 enters information such as "An invention of a drone holding an umbrella" into this text input field. Below the text input field, there are radio buttons for selecting a drawing type 33. The options available are "Block diagram," "Flowchart," "UI diagram," "Usage diagram," and "Leave it to AI." The drawing type acquisition unit 16 acquires the drawing type 33 when the user 20 selects one of these options. If the user 20 selects "Leave it to AI," the drawing type acquisition unit 16 estimates the drawing type 33 using an AI generation model based on the information received by the reception unit 11. The determination unit 12 determines which drawing creation method, the first image creation unit 13 or the second image creation unit 14, to use according to the acquired drawing type 33. After user 20 completes the input and selection, clicking the "Execute" button starts the processing by the patent drawing creation device 1. [Explanation of symbols]

[0053] 1. Patent drawing creation device 11 Reception Department 12 Judgment section 13. First Image Creation Section 14. Second Image Creation Section 15 Output section 16. Drawing Type Acquisition Section 20 users 31. Information regarding the drawings to be generated. 32 Patent Drawings 33 Types of Drawings

Claims

1. A reception unit that receives information regarding the drawings to be generated, A first image creation unit creates patent drawings using AI image generation technology based on the information received at the aforementioned reception unit, A second image creation unit creates a code for generating drawings based on the information received by the reception unit and creates a patent drawing from the said code. A determination unit determines, based on the information received by the reception unit, whether to use the drawing creation method of the first image creation unit or the second image creation unit, An output unit that outputs patent drawings created by the first image creation unit or the second image creation unit, The system includes a drawing type acquisition unit that acquires the type of drawing based on the information received at the aforementioned reception unit, The determination unit determines, according to the type of drawing, whether to use the drawing creation method of the first image creation unit or the second image creation unit. A patentable drawing creation device characterized by the following features.

2. A patent drawing creation apparatus according to claim 1, The first image creation unit has multiple drawing creation methods, The patented drawing creation apparatus is characterized in that the determination unit determines, based on the information received by the reception unit, whether to use one of the plurality of drawing creation methods of the first image creation unit or the drawing creation method of the second image creation unit.

3. A patent drawing creation apparatus according to claim 1, The second image creation unit has multiple drawing creation methods, The patented drawing creation apparatus is characterized in that the determination unit determines, based on the information received by the reception unit, whether to use the drawing creation method of the first image creation unit or one of the multiple drawing creation methods of the second image creation unit.

4. A patent drawing creation apparatus according to claim 1, The drawing type acquisition unit is characterized by acquiring the type of drawing based on input or selection from the user, or by estimating the type of drawing using an AI generation model based on information received by the reception unit.

5. A method for creating patent drawings using a computer, A receiving process for receiving information about the drawings to be generated, A first image creation step involves creating a patent drawing using AI image generation technology based on the information received in the aforementioned reception step, A second image creation step involves creating a code for generating drawings based on the information received in the aforementioned reception step, and creating patent drawings from said code. A determination step, based on the information received in the reception step, determines whether to use the drawing creation method of either the first image creation step or the second image creation step, An output step for outputting the patent drawing created in the first image creation step or the second image creation step, Based on the information received in the aforementioned reception process, a drawing type acquisition process is performed to acquire the drawing type, Includes, The determination step determines whether to use the drawing creation method of either the first image creation step or the second image creation step, depending on the type of drawing. A method for creating patent drawings characterized by the following features.

6. On the computer, A reception process that receives information about the drawings to be generated, A first image creation process that creates a patent drawing using AI image generation technology based on the information received in the aforementioned reception process, A second image creation process involves creating a code for generating drawings based on the information received in the aforementioned reception process, and creating a patent drawing from said code. A determination process that determines whether to use the drawing creation method of either the first image creation process or the second image creation process based on the information received in the aforementioned reception process, Output process for outputting patent drawings created in the first image creation process or the second image creation process, Based on the information received in the aforementioned reception process, a drawing type acquisition process is performed to acquire the drawing type. The determination process determines whether to use the drawing creation method of either the first image creation process or the second image creation process, depending on the type of drawing. A patent drawing creation program characterized by the following features.

Images