Modeling data conversion system, modeling data conversion method, and modeling data conversion program

Abstract

This reduces the workload for users to create measured objects such as human body parts using a 3D printer, which was previously complicated. [Solution] A model production data conversion system acquires 3D scan data of a measurement object and converts the acquired 3D scan data into 3D printer data for modeling the measurement object using a 3D printer. The model production data conversion system includes at least one control unit and a storage unit that stores a trained model using supervised data, where input data is 3D scan data acquired by a 3D scanner that scans the measurement object in three dimensions and before noise removal, and output data is processed data that has been noise removed and converted into data for the 3D printer. The control unit acquires the 3D scan data via a communications network and generates processed data from the 3D scan data using the trained model.

Description

[Technical Field]

[0001] The present invention relates to a model-making data conversion system, a model-making data conversion method, and a model-making data conversion program. [Background technology]

[0002] A variety of objects have been produced using 3D printers, including medical body part models. When producing a body part model using a 3D printer, the following process is typically followed:

[0003] (i) A doctor obtains data (DICOM data, etc.) of the target human body part through a CT scan. (ii) The CT scan data is transferred via media such as a CD-R, and the creator of the model manually removes various noises present in the CT scan data (for example, halation that occurs when X-rays are irradiated onto metal objects embedded in the body).

[0004] (iii) Using dedicated software, the worker manually converts the processed data after noise removal into 3D printer data (e.g., STL format) that can be read by a 3D printer. (iv) The modeling material to be used and the type of 3D printer to be used are selected depending on the object to be created. (v) The object is created using the 3D printer.

[0005] As a technology for acquiring data on human body parts and creating a model using a 3D printer, for example, Patent Document 1 discloses a 3D data generation device for a 3D printer that is equipped with a 3D data generation unit that receives 3D medical image data of the inner cavity of tubular tissue and generates 3D data to be used in the 3D printer. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Japanese Patent Publication No. 2024-067003 Summary of the Invention [Problem to be solved by the invention]

[0007] However, the process of manually removing various noises from CT scan data (for example, halation that occurs when X-rays are irradiated onto metal objects embedded in the body) and creating data that can be read by a 3D printer is extremely time-consuming and depends on the skill of the worker. As a result, these processes can incur significant costs, and variations in workmanship can occur, making it difficult to consistently create accurate data for 3D printers.

[0008] The present invention aims to provide a modeling data conversion system, a modeling data conversion method, and a modeling data conversion program that reduce the workload on users in the previously complicated process of producing measured objects such as human body parts using a 3D printer. [Means for solving the problem]

[0009] The object production data conversion system according to the invention described in claim 1 is a data conversion system for converting acquired 3D scan data of a measured object into 3D printer data for modeling the measured object using a 3D printer, and comprises at least one or more control units; and a memory unit that stores a trained model using supervised data, in which input data is 3D scan data acquired by a 3D scanner that scans the measured object in three dimensions and before noise removal, and output data is processed data from which noise has been removed and converted into data for a 3D printer, wherein the control unit acquires the 3D scan data via a communication network, and generates the processed data from the 3D scan data using the trained model.

[0010] The object production data conversion system according to the invention described in claim 2 is characterized in that the control unit uses the processed data to form the measured object using the 3D printer.

[0011] The object production data conversion system according to the invention described in claim 3 is characterized in that the control unit selects a modeling material or the 3D printer depending on the object to be measured.

[0012] The object production data conversion system according to the invention described in claim 4 is characterized in that the control unit calculates a cost estimate for the work of producing an object using the 3D printer based on the processed data.

[0013] In the object manufacturing data conversion system according to the invention described in claim 5, the noise includes halation or artifact noise.

[0014] The object production data conversion method of the invention described in claim 6 is a method for converting object production data using an object production data conversion system that acquires 3D scan data of a measured object and converts the acquired 3D scan data into 3D printer data for modeling the measured object with a 3D printer, characterized in that the object production data conversion system acquires the 3D scan data via a communications network, and generates the processed data from the acquired 3D scan data using a trained model based on supervised data, in which input data is 3D scan data acquired by a 3D scanner that scans the measured object in three dimensions and before noise removal, and output data is processed data from which noise has been removed and converted into 3D printer data.

[0015] The object production data conversion program of the invention described in claim 7 is a object production data conversion program executed by an object production data conversion system that acquires 3D scan data of a measured object and converts the acquired 3D scan data into 3D printer data for modeling the measured object with a 3D printer, and is characterized by including the steps of: acquiring 3D scan data via a communications network; and generating processed data from the acquired 3D scan data using a trained model based on supervised data, in which 3D scan data acquired by a 3D scanner that scans the measured object in three dimensions and before noise removal is used as input data, and processed data from which noise has been removed and converted into 3D printer data is used as output data. [Effects of the Invention]

[0016] According to the invention described in claim 1, it is possible to automate a series of processes that have been conventionally performed manually and are cumbersome, such as removing noise from 3D scan data and converting it into data that can be read by a 3D printer. This reduces the workload on the user in producing measured objects such as human body parts using a 3D printer.

[0017] According to the invention described in claim 2, the process from acquiring CT scan data to creating a model using a 3D printer can be performed in a consistent manner, eliminating the need for laborious processes such as considering and setting up modeling conditions.

[0018] According to the present invention as set forth in claim 3, it is possible to automatically select or suggest an appropriate material for the object to be measured and a 3D printer that is compatible with that material, thereby reducing the dependency on the user's skills and know-how required for modeling and also reducing the workload for the user performing the modeling.

[0019] According to the present invention as set forth in claim 4, since information on the estimated cost calculated using the trained model can be sent to the 3D data acquisition system, information on the cost required to produce the model can be quickly obtained from the provision of 3D scan data and provided to the person requesting the production of the model. This reduces the time required from 3D scanning the object to determining whether or not to produce the model.

[0020] According to the present invention as set forth in claim 5, it is possible to eliminate the effort required for manual noise removal, and to reduce the number of incorrect shapes being formed due to inaccurate unevenness or shapes during modeling.

[0021] According to the present invention as set forth in claims 6 and 7, it is possible to automate a series of processes that have conventionally been complicated, such as removing noise from 3D scan data and converting it into data that can be read by a 3D printer, thereby reducing the workload on the user in producing a measurement target such as a human body part using a 3D printer. [Brief explanation of the drawings]

[0022] [Figure 1] FIG. 1 is an overall configuration diagram of a model manufacturing system. [Figure 2] 10 is a flowchart of control of the object manufacturing system. DETAILED DESCRIPTION OF THE INVENTION

[0023] An embodiment of the present invention will be described below with reference to the drawings. Fig. 1 is a diagram showing the overall configuration of a model production system 1. The model production system 1 includes a data acquisition system 11 and a model production data conversion system 12. The data acquisition system 11 includes a 3D scanner 2 that measures a measurement target T and acquires 3D scan data D1.

[0024] The 3D scanner 2 is a device capable of acquiring 3D scan data D1 of a three-dimensional shape by three-dimensionally scanning a measurement object T. The 3D scanner 2 can use any method, such as a computed tomography scan (CT scan), magnetic resonance imaging (MRI), ultrasound scan, time-of-flight (ToF), a LiDAR sensor, laser triangulation, structured light surveying, stereo camera photography, structure-from-motion / multi-view-stereo (SfM-MVS), contact scanning using a probe, a phase shift method, etc.

[0025] Furthermore, the 3D scanner 2 can be any device depending on the size or scale of the measurement object T, and can be, for example, a handheld type, a tabletop type, a stand type, or an aerial type (e.g., a drone, an aircraft).

[0026] The measurement object T may be, for example, a sealant, a waterproof material, a gasket, a cover, or a medical surgical training model, or may be a building material part, a home appliance part, or a living body part, such as an internal organ or bone, which is acquired by a CT scanner or MRI.

[0027] The object production data conversion system 12 acquires 3D scan data D1 of the measurement object T, and converts the acquired 3D scan data D1 into 3D printer data for molding the measurement object T using a 3D printer 4.

[0028] The object production data conversion system 12 includes one or more devices. The object production data conversion system 12 is configured as, for example, a cloud server.

[0029] In this embodiment, a model-making data conversion device 3 is used as an example of a device constituting the model-making data conversion system 12, but some or more of the functional units of the model-making data conversion device 3 (control unit 31, communication unit 32, input unit 33, output unit 34, memory unit 35, etc.) may be provided in or distributed among multiple devices, and may function by the devices working together.

[0030] The object production data conversion device 3 is assumed to be a server, but other types of devices such as a portable or desktop personal computer, a mobile phone, a smartphone, or a tablet can also be used.

[0031] The object production data conversion device 3 includes at least one or more control units 31, communication units 32, input units 33, output units , and storage units .

[0032] The control unit 31 is a processor, and may be, for example, a central processing unit (CPU), a micro-processing unit (MPU), a GPU, a microcontroller unit (MCU), a processor core, a multiprocessor, an ASIC, an FPGA, or the like.

[0033] The control unit 31 may implement each process disclosed in the embodiments by a logic circuit formed in an integrated circuit or a dedicated circuit.

[0034] The communication unit 32 has a function of establishing a wired or wireless communication connection with an external device or system via a network such as a WAN or a LAN.

[0035] The input unit 33 has a function of inputting information and instructions to the program of the object production data conversion device 3. The input unit 33 may also have a function of detecting or receiving input of other information, such as a physical switch that receives instructions from a user, a sound collection unit, a light receiving unit, an imaging unit (camera), an acceleration sensor, or the like.

[0036] The output unit 34 has a function of outputting information to the outside of the object production data conversion device 3. The output unit 34 includes, for example, a display unit and a sound emitting unit. The display unit can be a display device equipped with a display screen such as a liquid crystal or OLED.

[0037] The storage unit 35 stores a modeling data conversion program 351, 3D scan data 353 acquired from the data acquisition system 11, processed data 354, modeling conditions 355 for modeling using the processed data 354, and estimate data 356.

[0038] The storage unit 35 may store other programs in addition to the object production data conversion program 351. The storage unit 35 that stores these programs can be realized by various non-transitory storage media such as an HDD, an SSD, or a flash memory.

[0039] The object production data conversion program 351 is a program that converts 3D scan data 353 of the measurement object T acquired from the data acquisition system 11 into processed data 354, which is 3D printer data for manufacturing the measurement object T using a 3D printer. The object production data conversion program 351 is executed by the object production data conversion system 12.

[0040] The object production data conversion program 351 includes a trained model 352. The trained model 352 is an AI (artificial intelligence) model (program) that has been trained in advance using supervised data, and that uses 3D scan data 353 before noise removal, acquired by a 3D scanner 2 that three-dimensionally scans the measurement target T, as input data, and that uses processed data, from which noise has been removed and converted into data for a 3D printer, as output data.

[0041] In this embodiment, the output data of the supervised data includes modeling conditions and estimate data. The modeling conditions can include the type of measurement object T (a specific object among building material parts, home appliance parts, or biological parts), the type of 3D printer (details will be described later), and modeling material.

[0042] The trained model 352 can be constructed using, for example, a convolution neural network (CNN), a recurrent neural network (RNN), a long short-term memory (LSTM), or the like.

[0043] The 3D scan data 353 is data in a format such as DICOM, acquired by the 3D scanner 2 of the data acquisition system 11 via the communication unit 32. For example, if the 3D scanner 2 is a CT scanner, the 3D scan data 353 includes a tomographic image (tif format) and a Molcer file (mol format).

[0044] The processed data 354 is data obtained by converting the 3D scan data 353 using the object production data conversion program 351 into a data format (for example, STL format) that can be read and processed by the 3D printer 4.

[0045] The modeling conditions 355 include the type of measurement object T (a specific object among building material parts, home appliance parts, or biological parts), the type of 3D printer, modeling materials, and modeling parameters.

[0046] The type of 3D printer can be one or more selected from a modeling method such as MEX (fused deposition modeling) or stereolithography (SLA / DLP), or from a number of pre-registered devices (3D printers managed or owned by businesses using the model production data conversion system 12 or other external businesses).

[0047] The modeling material is selected according to the type of 3D printer suitable for modeling the processed data 354, and may include, for example, resin, metal, ceramics, glass, or a composite of these.

[0048] Specifically, the material may include one or more of PLA (polylactic acid), silicone, ABS (acrylonitrile butadiene styrene), PETG (polyethylene terephthalate glycol), TPU (thermoplastic polyurethane), nylon, carbon fiber composite, nylon + glass beads, TPU powder, stainless steel, aluminum alloy, titanium alloy, Inconel (nickel alloy), wood filament (PLA + wood powder), conductive filament, water-soluble support material (PVA, HIPS), etc.

[0049] The modeling parameters may include one or more of the following: layer height, nozzle diameter, printing speed, extrusion temperature, bed temperature, infill rate, infill pattern (honeycomb or grid, etc.), support structure, environmental temperature control, and photolithography exposure time.

[0050] The estimate data 356 is data regarding the service cost that the business operator that manages or owns the 3D printer 4 presents to the client (for example, the business operator of the data acquisition system 11 that provides the 3D scan data 353) when a model is created using the 3D printer 4 using the processed data 354.

[0051] The estimate data 356 is calculated based on the selected modeling conditions 355, for example.

[0052] Next, each step of the flowchart of the control method or operation method of the object production system 1 using the object production data conversion program 351 will be described with reference to Fig. 2. The object production system 1 can execute the object production data conversion method using the object production data conversion system 12.

[0053] In step S01, the data acquisition system 11 uses the 3D scanner 2 to three-dimensionally scan and measure the measurement object T, and acquires 3D scan data.

[0054] In step S02, the control unit 31 acquires the 3D scan data D1 from the data acquisition system 11 via the communication network, and stores the data in the storage unit 35 as 3D scan data 353.

[0055] The measured 3D scan data 353 may contain noise such as halation or artifact noise, and may not be usable as data for a 3D printer as is. Therefore, in step S03, the control unit 31 uses the trained model 352 to remove noise from the 3D scan data 353 acquired by the molded object production data conversion system 12 (molded object production data conversion device 3).

[0056] The control unit 31 further converts the data into a data format (for example, STL format) that can be read and processed by the 3D printer 4, and generates processed data 354. The processed data 354 is stored in the storage unit 35.

[0057] In step S04, the control unit 31 uses the trained model 352 of the object production data conversion program 351 to output the modeling conditions for forming an object of the processed data 354 by the 3D printer 4. The output modeling conditions are stored in the memory unit 35 as modeling conditions 355.

[0058] In addition, the object production data conversion system 12 stores information about commercially available 3D printers in the memory unit 35, and in step S04, the control unit 31 may identify or recommend an appropriate 3D printer from the memory unit 35 depending on the object to be created.

[0059] As mentioned above, the object production data conversion device 3 also stores information about the main types of 3D printers. Because the 3D printer industry is developing rapidly, the device may be equipped with a function to add and update information on new modeling conditions, including 3D printer models, as needed.

[0060] Furthermore, the control unit 31 can use the trained model 352, which has learned from past data and technical documents, to present options such as, if a flexible measurement object T such as an organ is to be formed, "it is best to use a 3D printer △△ using silicone as the material and a material with hardness Hs of 'XX'," or, if a hard object such as a bone with a size of XX cm x XX cm x XX cm is to be formed, "it is best to use the rigid 3D printer '△△' by ●● company, use □□ as the material, and use a filling rate of XX%."

[0061] In step S05, the control unit 31 selects or determines a modeling material or a 3D printer 4 according to the measurement target. The modeling material or the 3D printer 4 may be selected or determined by a user's instruction. Note that the modeling material or the 3D printer 4 may be selected or determined when actually performing modeling, rather than at the stage of step S05.

[0062] By approving or selecting the modeling conditions output via the output unit 34 or the recommended modeling conditions, the user can select the device or modeling conditions to provide the modeling service from a wider range of options.

[0063] In step S06, the control unit 31 calculates, based on the processed data, a cost estimate for the work of forming an object using the 3D printer 4. The control unit 31 stores the calculated estimate data D2 in the storage unit 35 as estimate data 356.

[0064] In step S07, the control unit 31 transmits the estimate data 356 (D2) to the data acquisition system 11 or a device managed or owned by the user of the data acquisition system 11. This allows the user who provided the 3D scan data D1 to understand the service fee required to model the measurement object, and to determine whether or not to perform modeling using the modeling service and the number of items to be modeled.

[0065] In step S08, the control unit 31 uses the processed data 354 to model the measurement object using the 3D printer 4. When providing a modeling service based on the 3D scan data D1 provided by another company, the modeling of the object by the 3D printer 4 may be performed after receiving a modeling instruction from the user in step S07.

[0066] In the above, in this embodiment, the control unit 31 acquires 3D scan data D1 via a communication network and generates processed data 354 from the 3D scan data 353 using a trained model 352. This describes the object production data conversion system 12, the object production data conversion method, and the object production data conversion program 351.

[0067] This automates the previously tedious manual process of removing noise from 3D scan data and converting it into data that can be read by a 3D printer, reducing the workload for users in the process of producing measured objects such as human body parts using a 3D printer.

[0068] In this embodiment, the trained model 352 is illustrated as being part of the object production data conversion program 361, but the trained model 352 may also be executed by calling the trained model 352 as an external program.

[0069] Furthermore, in this embodiment, an example has been described in which the object production data conversion system 12 processes steps S04 to S06 using the object production data conversion program 351, but the object production data conversion system 12 may execute some or all of steps S04 to S06 using a program other than the object production data conversion program 351.

[0070] Furthermore, the trained model that executes (infers) steps S04 to S06 may be configured by one trained model 352 as described in this embodiment, or may be a different model, and is not limited to the above embodiment. [Explanation of symbols]

[0071] 1. Modeling system 2. 3D scanner 3. Modeling data conversion device 4. 3D printer 11 Data Acquisition System 12. Modeling data conversion system 31 Control Unit 32 Communications Department 33 Input section 34 Output section 35 Storage section 351 Modeling data conversion program 352 trained models 353 3D scan data 354 Processed Data 355 Modeling Conditions 356 Estimate Data D1 3D scan data D2 Estimate data T Measurement object

Claims

1. A model production data conversion system that converts acquired 3D scan data of a measurement object into 3D printer data for modeling by a 3D printer, At least one control unit; A memory unit that stores a trained model using supervised data, in which the 3D scan data obtained by a 3D scanner that three-dimensionally scans the measurement object and before noise removal is used as input data, and processed data from which noise has been removed and converted into data for the 3D printer is used as output data. The control unit acquires 3D scan data via a communication network, generating the processed data from the 3D scan data using the trained model; A modeling data conversion system characterized by:

2. The object production data conversion system according to claim 1 , wherein the control unit uses the processed data to model the measurement object using the 3D printer.

3. The object production data conversion system according to claim 2 , wherein the control unit selects a modeling material or the 3D printer depending on the object to be measured.

4. The object production data conversion system according to claim 1 , wherein the control unit calculates a cost estimate for an operation of producing an object using the 3D printer based on the processed data.

5. 2. The system for converting data for producing a molded object according to claim 1, wherein the noise includes halation or artifact noise.

6. A model production data conversion method using a model production data conversion system that acquires 3D scan data of a measurement object and converts the acquired 3D scan data into 3D printer data for modeling the measurement object with a 3D printer, The object production data conversion system acquires 3D scan data via a communication network, The processed data is generated from the acquired 3D scan data by the object production data conversion system using a trained model based on supervised data, in which 3D scan data before noise removal is acquired by a 3D scanner that three-dimensionally scans the measurement object and 3D scan data before noise removal is used as input data, and processed data from which noise has been removed and converted into data for a 3D printer is used as output data. A method for converting data for producing a molded object, comprising:

7. A modeling data conversion program executed by a modeling data conversion system that acquires 3D scan data of a measurement object and converts the acquired 3D scan data into 3D printer data for modeling the measurement object by a 3D printer, acquiring 3D scan data via a communications network; A process of generating processed data from the acquired 3D scan data using a trained model based on supervised data, in which 3D scan data obtained by a 3D scanner that three-dimensionally scans the measurement object and before noise removal is used as input data, and processed data from which noise has been removed and converted into data for a 3D printer is used as output data; 10. A modeling data conversion program comprising:

Images