Three-dimensional shape data creation method, program, and three-dimensional shape data creation device

Abstract

A three-dimensional shape data creation method and the like are provided that can suppress the dependence of the review process on individual personnel and the increase in developer work costs. [Solution] The three-dimensional shape data creation method includes an acquisition unit (11) acquiring first three-dimensional shape data for creating a target molded product, a conversion unit (12) converting the first three-dimensional shape data into second three-dimensional shape data which is a voxel set defined by at least one or more voxels, a basic point set defined by at least one or more points, or a point voxel set defined by at least one or more points and voxels, and a generation unit generating molded product three-dimensional shape data which represents the shape of the target molded product using a machine learning model which has been trained in advance by machine learning, using the second three-dimensional shape data as input.

Description

[Technical Field]

[0001] The present disclosure relates to a three-dimensional shape data creation method, a program, and a three-dimensional shape data creation device. [Background technology]

[0002] For example, Patent Document 1 discloses a mold apparatus that can injection-mold a molded product using a first mold, a second mold, and a nest. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2023-152777 Summary of the Invention [Problem to be solved by the invention]

[0004] For example, a manufacturer may manufacture a molded product at the request of a customer. In this case, developers receive input data such as blueprints or 3D data, as well as customer requirements such as the appearance of the molded product, gate placement, areas that cannot be changed, and materials. Based on this input, developers go through a process of examining the mold structure, recessed area, and resin flow required to manufacture the molded product, then model a draft and verify the moldability risks of the modeled draft. After verification, developers then propose a molded product shape model to the customer.

[0005] However, because the review process is carried out based on the developer's internal thinking, it can become highly individualized. Also, because the developer must model the draft and verify the manufacturability risks of the modeled draft, the developer's work costs increase.

[0006] In Patent Document 1, a molded product is injection molded using a first mold, a second mold, and a nest, but no measures are taken to prevent the above-mentioned review process from becoming dependent on individual skills and to prevent an increase in developer work costs.

[0007] The present disclosure aims to provide a three-dimensional shape data creation method and the like that can suppress the dependency of the review process on individual tasks and the increase in developer work costs. [Means for solving the problem]

[0008] The three-dimensional shape data creation method according to the present disclosure includes: an acquisition unit acquiring first three-dimensional shape data for creating a target molded product; and a conversion unit converting the first three-dimensional shape data into second three-dimensional shape data, which is a voxel set defined by at least one or more voxels, a basic point set defined by at least one or more points, or a point voxel set defined by at least one or more points and voxels; and inputting the second three-dimensional shape data, A method for expressing the shape of the molded product from the second three-dimensional shape data The method includes a generation unit generating three-dimensional shape data of a molded product that represents the shape of a target molded product using a machine learning model that has been trained in advance by machine learning. The first three-dimensional shape data is data that represents a three-dimensional representation of the molded product having at least one of a flow path shape for material movement and a fastening hole for passing a fastening part. . [Effects of the Invention]

[0009] According to the three-dimensional shape data creation method and the like disclosed herein, it is possible to suppress the dependence of the review process on individual people and the increase in developer work costs. [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 1 is a block diagram showing a three-dimensional shape data generating device according to an embodiment. [Figure 2] FIG. 2 is a diagram showing the process from the first three-dimensional shape data to the second three-dimensional shape data of the molded product. [Figure 3] FIG. 3 is a diagram illustrating the learning process of the first machine learning model. [Figure 4] FIG. 4 is a diagram illustrating the learning process of the second machine learning model. [Figure 5] FIG. 5 is a flowchart showing an example of the operation of the three-dimensional shape data creation device according to the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. Note that each of the embodiments described below represents a specific example of the present disclosure. Therefore, the numerical values, shapes, materials, components, component arrangements and connection forms, steps, step order, etc. shown in the following embodiments are merely examples and are not intended to limit the present disclosure. Therefore, among the components in the following embodiments, components not recited in independent claims will be described as optional components.

[0012] Furthermore, each figure is a schematic diagram and is not necessarily an exact illustration. Therefore, for example, the scales and the like do not necessarily match in each figure. Furthermore, in each figure, substantially the same configurations are assigned the same reference numerals, and duplicate explanations are omitted or simplified.

[0013] Hereinafter, the embodiments will be specifically described with reference to the drawings.

[0014] (Embodiment) <Configuration and Function> First, the configuration and functions of a three-dimensional shape data creation device 1 according to this embodiment will be described with reference to FIGS.

[0015] FIG. 1 is a block diagram showing a three-dimensional shape data creation device 1 according to an embodiment. FIG. 2 is a diagram showing the process from first three-dimensional shape data to second molded product three-dimensional shape data. (a) of FIG. 2 shows the first three-dimensional shape data expressed in STL format, (b) of FIG. 2 shows the second three-dimensional shape data expressed in voxel format, (c) of FIG. 2 shows the first machine learning model, (d) of FIG. 2 shows the first molded product three-dimensional shape data, (e) of FIG. 2 shows the second machine learning model, and (f) of FIG. 2 shows the second molded product three-dimensional shape data. FIG. 3 is a diagram showing the learning process of the first machine learning model. FIG. 4 is a diagram showing the learning process of the second machine learning model.

[0016] As shown in FIGS. 1 and 2, the three-dimensional shape data creation device 1 includes an acquisition unit 11, a conversion unit 12, a generation unit, and a storage unit.

[0017] The acquisition unit 11 is configured to acquire first three-dimensional shape data for creating a target molded article (product). The first three-dimensional shape data is data that represents a three-dimensional representation of the target surface shape (including contours and ridges on the surface), and is a collection of coordinate data that expresses the positions of measurement points on the target surface using an XYZ coordinate system. The target of this embodiment is, for example, the shape of a flow path through which a substance moves inside a product. The flow path shape is a three-dimensional representation of the shape of a passage through which a liquid, gas, powder, or other substance flows. The first three-dimensional shape data shown in FIG. 2(a) is an example of three-dimensional shape data in STL format that represents the flow path shape.

[0018] The conversion unit 12 is configured to convert the first three-dimensional shape data acquired by the acquisition unit 11 into second three-dimensional shape data. The second three-dimensional shape data is a voxel set defined by at least one or more voxels, a basic point set defined by at least one or more points, or a point voxel set defined by at least one or more points and voxels. For example, as shown in FIG. 2(b), the conversion unit 12 converts the first three-dimensional shape data in STL format representing the flow path shape into second three-dimensional shape data representing the flow path shape in voxel format.

[0019] The generation unit is configured to receive the second three-dimensional shape data as input and generate three-dimensional molded product shape data representing the shape of the target molded product using a trained machine learning model that has been trained in advance by machine learning.

[0020] The machine learning model includes a first machine learning model and a second machine learning model. The molded product three-dimensional shape data includes first molded product three-dimensional shape data representing the shape of a first molded product, which is a precursor to a target molded product, and second molded product three-dimensional shape data representing the shape of a second molded product, which is the target molded product.

[0021] Specifically, the generation unit includes a first generation unit 21 and a second generation unit 22. The storage unit includes a first storage unit 31 and a second storage unit 32.

[0022] As shown in Figures 1 and 2 (b), (c), and (d), the first generation unit 21 is configured to receive the second three-dimensional shape data as input and generate the first molded product three-dimensional shape data using the first machine learning model.

[0023] The first machine learning model is a machine learning model that takes into account the flow path shape, and the wall thickness, ribs, and undercuts of the molded article that form the flow path shape when forming the first molded article. Note that the machine learning model may further include a model that takes into account fastening holes and the wall thickness, ribs, and undercuts of the molded article that form the fastening holes. The fastening holes are holes for fastening with fastening members such as bolts.

[0024] The first molded product three-dimensional shape data is data that represents the surface shape of the first molded product (including the contour and ridges on the surface) in three dimensions, and is a collection of coordinate data that expresses the positions of measurement points on the surface of the first molded product using an XYZ coordinate system. As shown in Fig. 2(d), the first molded product three-dimensional shape data of this embodiment represents the first molded product having a flow path shape in three dimensions. The first molded product three-dimensional shape data may also represent the first molded product having fastening holes in three dimensions.

[0025] The first molded product is a preliminary molded product having a shape that represents the target molded product, and is a molded product in which at least the flow path shape is formed by the molded product wall. The first molded product further has ribs and undercuts formed therein. The first molded product does not have gates connected to cavities for forming the target molded product, connectors for connecting to the target molded product, etc.

[0026] As shown in (d), (e), and (f) of Figures 1 and 2, the second generation unit 22 is configured to receive the first molded product three-dimensional shape data as input and generate the second molded product three-dimensional shape data using the second machine learning model.

[0027] The second machine learning model is a machine learning model that takes into account a gate connected to a cavity for forming a target molded product (second molded product) when molding the target molded product using a mold. Furthermore, the second machine learning model may be a machine learning model that takes into account a connecting portion for connecting to the second molded product when molding the target molded product (second molded product). The connecting portion is, for example, a connecting portion for connecting to an object.

[0028] The second molded product three-dimensional shape data is data that represents the surface shape of the second molded product (including the contour and ridges on the surface) in three dimensions, and is a collection of coordinate data that expresses the positions of measurement points on the surface of the second molded product using an XYZ coordinate system. As shown in Figure 2 (f), the second molded product three-dimensional shape data of this embodiment represents the second molded product having a flow path shape in three dimensions. The second molded product three-dimensional shape data may also represent the first molded product having fastening holes in three dimensions.

[0029] The second molded article represents the shape of the target molded article, and is a molded article formed with a gate connected to a cavity for forming the target molded article in addition to the first molded article. The second molded article may further have a connecting portion formed therein.

[0030] The second generating unit 22 outputs the generated second molded product three-dimensional shape data. For example, the second molded product three-dimensional shape data may be three-dimensional shape data in a voxel format that represents the shape of the product (second molded product).

[0031] The first storage unit 31 is configured with a semiconductor memory or the like, and stores the first machine learning model, the first molded product three-dimensional shape data, etc. The first storage unit 31 may be a semiconductor memory mounted in the first generation unit 21.

[0032] The second storage unit 32 is configured with a semiconductor memory or the like, and stores the second machine learning model, the second three-dimensional shape data of the molded product, etc. The second storage unit 32 may be a semiconductor memory mounted in the second generation unit 22.

[0033] Next, the learning of the first machine learning model will be described.

[0034] As shown in FIG. 3, the first machine learning model is trained using a first training unit 41a and a first identification unit 41b.

[0035] The first learning unit 41a is a convolutional neural network architecture such as U-Net. The first learning unit 41a is composed of an encoder, a decoder, etc., and can achieve high-resolution output while retaining three-dimensional shape data.

[0036] The first learning unit 41a is not limited to U-Net, and may be, for example, a CNN architecture applicable to an encoder-decoder structure such as Attention U-Net, Residual U-Net, SegNet, Tiramisu Network (FC-DenseNet), 3DU-Net, etc. Alternatively, the first learning unit 41a may use other known architectures.

[0037] The first learning unit 41a receives the second three-dimensional shape data as input data and can generate first predicted three-dimensional shape data expressed in voxel format using the first machine learning model currently being trained.

[0038] The first identification unit 41b is a classifier such as Patch GAN used in image generation and conversion tasks.

[0039] The first discrimination unit 41b is not limited to Patch GAN, and may be, for example, a discrimination architecture that focuses on a local region, such as a Multi-Scale Discriminator, a Dual Discriminator, or Patch-based Discriminators with Overlapping Patches. Alternatively, the first discrimination unit 41b may use other known architectures.

[0040] The first discrimination unit 41b receives the first predicted three-dimensional shape data output by the first learning unit 41a and a true value representing the correct first molded product as input data, and determines whether the first predicted three-dimensional shape data is true or false. That is, the first discrimination unit 41b compares the first predicted three-dimensional shape data with the true value and determines whether the first predicted three-dimensional shape data matches the true value. The first discrimination unit 41b is capable of evaluating the local quality of an output image, such as three-dimensional shape data, and is able to determine the true or false for each small patch that constitutes the three-dimensional shape data and determine the true or false of the entire three-dimensional shape data based on the combination of patch results.

[0041] As a result, the first learning unit 41a can learn by calculating a loss function obtained from the results determined by the first identification unit 41b and updating the weighting so as to minimize the loss. This update is performed, for example, using the backpropagation method and based on an optimization algorithm such as SGD (Stochastic Gradient Descent). This allows the first learning unit 41a to generate first predicted 3D shape data that is identical to the true value.

[0042] For example, in the first learning unit 41a of this embodiment, the first machine learning model is trained so as to take into account the flow path shape, and the thickness of the molded product wall that forms the flow path shape, ribs, and undercuts.

[0043] For example, the training of the first machine learning model includes conditions for setting the molded product wall portion for forming the flow path shape and conditions for setting the wall thickness of the molded product. Specifically, the first training unit 41a and the first identification unit 41b can incorporate into the loss function, as constraints, key functional points such as the flow path shape, appearance requirements for the molded product, and positional information of parts whose shape cannot be changed. Therefore, the first training unit 41a can train the first machine learning model taking these constraints into account.

[0044] For example, in training the first machine learning model, the position where the rib is to be left can be added as a constraint. Specifically, the first training unit 41a and the first identification unit 41b can incorporate the position where the rib is to be left (rib position information) as a constraint into the loss function. Therefore, the first training unit 41a can train the first machine learning model taking the constraint into consideration.

[0045] As described above, the first generation unit 21 can generate first molded product three-dimensional shape data based on the input second three-dimensional shape data by using the first machine learning model learned by the first learning unit 41a.

[0046] Next, the learning of the second machine learning model will be explained.

[0047] As shown in FIG. 4, the second machine learning model is trained using a second training unit 42a and a second identification unit 42b.

[0048] The second learning unit 42a is a convolutional neural network architecture such as U-Net. The second learning unit 42a is composed of an encoder, a decoder, etc., and can achieve high-resolution output while retaining three-dimensional shape data.

[0049] The second learning unit 42a is not limited to U-Net, and may be, for example, a CNN architecture applicable to an encoder-decoder structure such as Attention U-Net, Residual U-Net, SegNet, Tiramisu Network (FC-DenseNet), 3DU-Net, etc. Alternatively, the second learning unit 42a may use other known architectures.

[0050] The second learning unit 42a can use the first molded product three-dimensional shape data as input data and generate second predicted three-dimensional shape data expressed in voxel format using the second machine learning model currently being learned.

[0051] The second identification unit 42b is a classifier such as Patch GAN used in image generation and conversion tasks.

[0052] The second identification unit 42b is not limited to Patch GAN, and may be, for example, a discrimination architecture that focuses on a local region, such as a Multi-Scale Discriminator, a Dual Discriminator, or Patch-based Discriminators with Overlapping Patches. Alternatively, the second identification unit 42b may use other known architectures.

[0053] The second discrimination unit 42b receives the second predicted three-dimensional shape data output by the second learning unit 42a and a true value representing the correct second molded product as input data, and determines whether the second predicted three-dimensional shape data is true or false. That is, the second discrimination unit 42b compares the second predicted three-dimensional shape data with the true value and determines whether the second predicted three-dimensional shape data matches the true value. The second discrimination unit 42b can evaluate the local quality of an output image, such as three-dimensional shape data, and can determine the true or false of each small patch that makes up the three-dimensional shape data and determine the true or false of the entire three-dimensional shape data based on the combination of patch results.

[0054] As a result, the second learning unit 42a can learn by calculating a loss function obtained from the results determined by the second identification unit 42b and updating the weighting so as to minimize the loss. This update is performed, for example, using the backpropagation method and based on an optimization algorithm such as SGD (Stochastic Gradient Descent). This allows the second learning unit 42a to generate second predicted 3D shape data that is identical to the true value.

[0055] For example, the second learning unit 42a in this embodiment generates the gate position while taking into account the opening and closing parts of the mold. The second learning unit 42a trains the second machine learning model so that, when molding a target molded product using a mold, it can take into account the gate connected to the cavity for forming the target molded product. Furthermore, the second learning unit 42a trains the second machine learning model so that, when molding the target molded product, it can take into account the connecting parts for connecting to the target molded product.

[0056] For example, in training the second machine learning model, positions where gates cannot be placed are added as conditions. Specifically, the second learning unit 42a and the second identification unit 42b can incorporate positions where gates cannot be placed (information on positions where gates cannot be placed) as constraints into the loss function. Therefore, the second learning unit 42a can train the second machine learning model taking these constraints into account.

[0057] For example, in training the second machine learning model, positions where connectors cannot be placed can be added to the constraints. Specifically, the second learning unit 42a and the second identification unit 42b can incorporate the positions where connectors cannot be placed (information on positions where gates cannot be placed) as constraints into the loss function. Therefore, the second learning unit 42a can train the second machine learning model taking the constraints into account.

[0058] In this way, the second generation unit 22 can generate second molded product three-dimensional shape data based on the input first molded product three-dimensional shape data by using the second machine learning model learned by the second learning unit 42a.

[0059] <Operation> First, the operation of the three-dimensional shape data creation device 1 in this embodiment will be described with reference to FIG.

[0060] FIG. 5 is a flowchart showing an example of the operation of the three-dimensional shape data creation device 1 according to the embodiment.

[0061] At the request of a customer, the developer extracts first three-dimensional shape data from product data such as drawing data or three-dimensional data. That is, the developer generates the first three-dimensional shape data by extracting essential shapes such as flow path shapes based on the customer's product data. The developer inputs the extracted first three-dimensional shape data to the three-dimensional shape data creation device 1, and the acquisition unit 11 can acquire the first three-dimensional shape data (S11).

[0062] Next, the conversion unit 12 converts the first three-dimensional shape data acquired by the acquisition unit 11 into second three-dimensional shape data (S12). For example, the conversion unit 12 converts the first three-dimensional shape data into second three-dimensional shape data that is a voxel set defined by at least one or more voxels, a basic point set defined by at least one or more points, or a point voxel set defined by at least one or more points and voxels, so that the first machine learning model can easily generate the first three-dimensional shape data of the molded product.

[0063] Next, the first generation unit 21 receives the second three-dimensional shape data as an input and generates first molded article three-dimensional shape data using the first machine learning model (S13). For example, the first generation unit 21 uses the first machine learning model to generate first molded article three-dimensional shape data in which at least the flow path shape is formed, taking undercuts into consideration, based on the second three-dimensional shape data in voxel format.

[0064] Next, the second generation unit 22 receives the first molded product three-dimensional shape data as input and generates second molded product three-dimensional shape data using the second machine learning model (S14). For example, the second generation unit 22 uses the first machine learning model to generate second molded product three-dimensional shape data in which a gate connected to a cavity for forming a target molded product is formed, based on the first molded product three-dimensional shape data in voxel format. In this manner, in this operation example, the second molded product three-dimensional shape data representing the target molded product (second molded product) can be obtained.

[0065] <Action and effect> Next, the effects of the three-dimensional shape data creation method, program, and three-dimensional shape data creation device 1 according to this embodiment will be described.

[0066] The three-dimensional shape data creation method of Technology 1 in this embodiment includes the steps of: an acquisition unit 11 acquiring first three-dimensional shape data for creating a target molded product; a conversion unit 12 converting the first three-dimensional shape data into second three-dimensional shape data which is a voxel set defined by at least one or more voxels, a basic point set defined by at least one or more points, or a point voxel set defined by at least one or more points and voxels; and a generation unit generating molded product three-dimensional shape data which represents the shape of the target molded product using a machine learning model which has been trained in advance by machine learning, using the second three-dimensional shape data as an input.

[0067] According to this, by inputting the second three-dimensional shape data to the generation unit, the machine learning model can generate three-dimensional shape data of the molded product.

[0068] Therefore, it is possible to prevent the review process from becoming dependent on a specific individual and to prevent an increase in developer work costs.

[0069] A three-dimensional shape data creation method of Technology 2 in this embodiment is the three-dimensional shape data creation method described in Technology 1. In this case, the machine learning model includes a first machine learning model and a second machine learning model, the molded product three-dimensional shape data includes first molded product three-dimensional shape data that is a precursor to the shape representing the target molded product, and second molded product three-dimensional shape data that represents the shape of the target molded product, and the generation unit includes a first generation unit 21 and a second generation unit 22, the first generation unit 21 receives the second three-dimensional shape data as an input and generates the first molded product three-dimensional shape data using the first machine learning model, and the second generation unit 22 receives the first molded product three-dimensional shape data as an input and generates the second molded product three-dimensional shape data using the second machine learning model.

[0070] This allows the generation of the first molded product three-dimensional shape data, which is the previous stage, after which the second molded product three-dimensional shape data, which will become the final product, can be generated, thereby reducing the increase in processing load on the generation unit compared to when the second molded product three-dimensional shape data is generated all at once.

[0071] For example, when a gate is included in the molded product three-dimensional shape data, the correct answer may be obtained even if the gate is formed in a different location. When multiple correct answer candidates exist, the accuracy of the molded product three-dimensional shape data may be reduced even if the molded product three-dimensional shape data is generated using a machine learning model. However, in this embodiment, by generating first molded product three-dimensional shape data, which is a precursor to the target molded product, and then generating second molded product three-dimensional shape data, the accuracy of the second molded product three-dimensional shape data representing the target molded product can be reduced.

[0072] The three-dimensional shape data creation method of Technology 3 in this embodiment is the three-dimensional shape data creation method described in Technology 1. In this case, the first three-dimensional shape data is data obtained by three-dimensionally representing the shape of a flow path through which a substance moves, and the molded product three-dimensional shape data is data obtained by three-dimensionally representing a target molded product having the shape of the flow path.

[0073] This makes it possible to obtain three-dimensional shape data of a molded product in which at least the flow path shape is formed.

[0074] The three-dimensional shape data creation method of Technology 4 in this embodiment is the three-dimensional shape data creation method described in Technology 3. In this case, the machine learning model is a machine learning model that takes into consideration the flow path shape, the wall thickness of the molded product that forms the flow path shape, ribs, and undercuts, and also takes into consideration the gate connected to the cavity for forming the target molded product using a mold, when molding the target molded product.

[0075] This makes it possible to generate one piece of 3D shape data for a molded product using a single machine learning model, taking into account the flow path shape, wall thickness of the molded product, ribs, and undercuts, as well as the gate connected to the cavity. This makes it possible to expect to obtain one piece of 3D shape data for molding a target molded product.

[0076] The three-dimensional shape data creation method of Technology 5 in this embodiment is the three-dimensional shape data creation method described in Technology 2. In this case, the first three-dimensional shape data is data obtained by three-dimensionally representing a flow path shape for substance movement, the first molded product three-dimensional shape data is data obtained by three-dimensionally representing a first molded product having the flow path shape, the first machine learning model is a machine learning model that takes into account the flow path shape and the thickness, ribs, and undercuts of the molded product wall that form the flow path shape, and the second machine learning model is a machine learning model that takes into account a gate connected to a cavity for forming the molded product when the target molded product is molded using a mold.

[0077] This makes it possible to generate first molded product three-dimensional shape data that takes into account the flow path shape, the molded product wall thickness, ribs, and undercuts using the first machine learning model, which is expected to enable obtaining second molded product three-dimensional shape data for molding a target molded product.

[0078] Furthermore, the second machine learning model makes it possible to generate second molded product three-dimensional shape data that takes gates into consideration, thereby obtaining second molded product three-dimensional shape data for molding a target molded product.

[0079] The three-dimensional shape data creation method of Technique 6 in this embodiment is the three-dimensional shape data creation method described in Technique 5. In this case, the second machine learning model is a machine learning model that takes into account a connection portion for connecting to a target molded product when molding the target molded product.

[0080] This makes it possible to generate second molded product three-dimensional shape data that takes into account gates and connecting parts using the second machine learning model, which is expected to enable obtaining second molded product three-dimensional shape data for molding a target molded product.

[0081] The program of Technique 7 in this embodiment is a program that enables a computer to execute the three-dimensional shape data creation method described in any one of Techniques 1 to 6.

[0082] This program also provides the same effects as those described above.

[0083] The three-dimensional shape data creation device 1 of Technology 8 in this embodiment includes an acquisition unit 11 that acquires first three-dimensional shape data for creating a target molded product, a conversion unit 12 that converts the first three-dimensional shape data into second three-dimensional shape data which is a voxel set defined by at least one or more voxels, a basic point set defined by at least one or more points, or a point voxel set defined by at least one or more points and voxels, and a generation unit that receives the second three-dimensional shape data as input and generates molded product three-dimensional shape data representing the shape of the target molded product using a machine learning model that has been trained in advance by machine learning.

[0084] This three-dimensional shape data creation device 1 also provides the same effects as those described above.

[0085] (Other variations) Although the three-dimensional shape data creation method and the like according to the present disclosure have been described based on the above-mentioned embodiments, the present disclosure is not limited to these embodiments. As long as they do not deviate from the spirit of the present disclosure, various modifications conceivable by those skilled in the art may also be included within the scope of the present disclosure.

[0086] For example, in the three-dimensional shape data creation method etc. according to the present disclosure, the generation unit may use one machine learning model to directly generate the second three-dimensional shape data of a molded product from the second three-dimensional shape data. In other words, one generation unit may be able to generate, using one machine learning model, three-dimensional shape data of a molded product corresponding to the second three-dimensional shape data of a molded product that represents the shape of a target molded product.

[0087] In this case, the machine learning model may be a model that takes into consideration the flow path shape, the wall thickness of the molded product that forms the flow path shape, ribs, and undercuts when molding the target molded product, and also takes into consideration a gate connected to a cavity for forming the target molded product using a mold. In this case, the generation unit may be configured to receive the second three-dimensional shape data as input and generate three-dimensional molded product shape data in voxel format that represents the shape of the product (target molded product) using the machine learning model.

[0088] Furthermore, in the three-dimensional shape data creation method and the like according to the present disclosure, molded articles having a flow path shape (e.g., valve bodies, joint pipes) are exemplified, but the molded articles are not limited to these. For example, the molded articles may be furniture such as chairs, desks, and tables, tableware such as cups, knives, and forks, tools such as wrenches and screwdrivers, mechanical parts such as gears, etc.

[0089] Furthermore, the generating unit and the like used in the 3D shape data creation method and the like according to the present disclosure are typically realized as an LSI, which is an integrated circuit. These may be individually implemented as single chips, or some or all of them may be integrated into a single chip.

[0090] Furthermore, the integration is not limited to LSI, but may be realized by dedicated circuits or general-purpose processors. FPGAs (Field Programmable Gate Arrays), which can be programmed after LSI fabrication, or reconfigurable processors, which allow the connections and settings of circuit cells within LSIs to be reconfigured, may also be used.

[0091] In each of the above embodiments, each component may be configured with dedicated hardware, or may be realized by executing a software program suitable for each component. Each component may also be realized by a program execution unit such as a CPU or processor reading and executing a software program recorded on a recording medium such as a hard disk or semiconductor memory.

[0092] Furthermore, all of the numbers used above are examples for specifically explaining the present disclosure, and the embodiments of the present disclosure are not limited to the numbers shown as examples.

[0093] The division of functional blocks in the block diagram is an example, and multiple functional blocks may be realized as a single functional block, one functional block may be divided into multiple blocks, or some functions may be moved to another functional block.Furthermore, the functions of multiple functional blocks having similar functions may be processed in parallel or in time-sharing by a single piece of hardware or software.

[0094] The order in which the steps in the flowchart are executed is merely an example for specifically explaining the present disclosure, and an order other than the above may be used. Also, some of the steps may be executed simultaneously (in parallel) with other steps.

[0095] In addition, this disclosure also includes forms obtained by making various modifications to the above embodiments that a person skilled in the art would think of, and forms realized by arbitrarily combining the components and functions of the embodiments within the scope that does not deviate from the intent of this disclosure. [Industrial Applicability]

[0096] The present disclosure is applicable to the manufacture of products using molds. [Explanation of symbols]

[0097] 1. Three-dimensional shape data creation device 11 Acquisition Department 12 Conversion unit 21 1st generation section (generation section) 22 Second generation section (generation section)

Claims

1. an acquisition unit acquires first three-dimensional shape data for creating a target molded product; A conversion unit converts the first three-dimensional shape data into second three-dimensional shape data which is a voxel set defined by at least one voxel, a basic point set defined by at least one point, or a point voxel set defined by at least one point and voxel; a generation unit generates molded product three-dimensional shape data representing a target shape of the molded product using a machine learning model that is trained in advance by machine learning to represent the shape of the molded product from the second three-dimensional shape data, using the second three-dimensional shape data as an input; The first three-dimensional shape data is data that represents the molded product in three dimensions, the molded product having at least one of a flow path shape for material movement and a fastening hole for passing a fastening part through. A method for creating three-dimensional shape data.

2. the machine learning models include a first machine learning model and a second machine learning model; the molded product three-dimensional shape data includes first molded product three-dimensional shape data, which is a preliminary stage of a shape representing a target molded product, and second molded product three-dimensional shape data representing the target shape of the molded product; the generating unit includes a first generating unit and a second generating unit; the first generation unit receives the second three-dimensional shape data as an input and generates the first molded product three-dimensional shape data using the first machine learning model; The second generation unit receives the first molded product three-dimensional shape data as an input and generates the second molded product three-dimensional shape data using the second machine learning model. The three-dimensional shape data creation method according to claim 1 .

3. the first three-dimensional shape data is data representing a three-dimensional shape of a flow path through which a substance moves, The molded product three-dimensional shape data is data that represents a target molded product having the flow path shape in three dimensions. The three-dimensional shape data creation method according to claim 1 .

4. The machine learning model is a machine learning model that takes into consideration the flow path shape, the wall thickness of the molded product that forms the flow path shape, ribs, and undercuts when molding the target molded product, and also takes into consideration a gate that is connected to a cavity for forming the target molded product using a mold. The three-dimensional shape data creation method according to claim 3.

5. the first three-dimensional shape data is data representing a three-dimensional shape of a flow path through which a substance moves, the first molded product three-dimensional shape data is data obtained by three-dimensionally representing a first molded product having the flow path shape, the first machine learning model is a machine learning model that takes into account the flow path shape, and a wall thickness, ribs, and undercuts of a molded product that form the flow path shape, The second machine learning model is a machine learning model that takes into account a gate connected to a cavity for forming the target molded product when the target molded product is molded using a mold. The three-dimensional shape data creation method according to claim 2.

6. The second machine learning model is a machine learning model that takes into account a connection portion for connecting to the target molded product when molding the target molded product.

6. The three-dimensional shape data creation method according to claim 5.

7. A method for creating three-dimensional shape data according to any one of claims 1 to 6, program.

8. an acquisition unit that acquires first three-dimensional shape data for creating a target molded product; a conversion unit that converts the first three-dimensional shape data into second three-dimensional shape data that is a voxel set defined by at least one voxel, a basic point set defined by at least one point, or a point voxel set defined by at least one point and voxel; a generation unit that receives the second three-dimensional shape data as an input and generates molded product three-dimensional shape data that represents a target shape of the molded product using a machine learning model that has been trained in advance by machine learning to represent the shape of the molded product from the second three-dimensional shape data, The first three-dimensional shape data is data that represents the molded product in three dimensions, the molded product having at least one of a flow path shape for material movement and a fastening hole for passing a fastening part through. Three-dimensional shape data creation device.

Images