AI model training device for collecting patterns of a printing substrate and inkjet printing apparatus

Abstract

The utility model relates to digital inkjet printing technical field, solve the existing printing material pattern in the change type adaptation process and consume the normal operation time of printing equipment, cause the problem of printing equipment's work cost is too high, provide the AI model training device and inkjet printing equipment of collection printing material pattern, AI model training device includes: camera for collecting the printing material image of the region to be printed, light source for providing illumination when the camera collects the printing material image of the region to be printed, model training unit is connected with the camera and receives the image data of printing material image of the region to be printed sent by the camera, the model training unit trains the image data to obtain the model data matching the region to be printed, the utility model has the advantage that for inkjet printing equipment can avoid wasting its normal operation time and significantly reduce its work cost.

Description

Technical Field

[0001] This utility model relates to the field of digital inkjet printing technology, and in particular to an AI model training device for acquiring patterns of a substrate and an inkjet printing device. Background Technology

[0002] Industrial inkjet printing has applications in many fields, such as digital printing, ceramics, printed circuit boards, mobile phone cases, and packaging boxes. Taking digital printing as an example, the pattern is input into the computer in digital form. The computer processes the pattern data to generate printing data that the printhead can recognize. The printing data controls the printhead to spray ink onto the substrate to form text and patterns, thus obtaining the printed product.

[0003] When dyeing substrates such as fabrics, traditional printing methods suffer from problems such as long printing time, high labor costs, and environmental pollution. With existing digital inkjet printing, the printing process requires changing the print pattern depending on the area to be printed on the substrate. This often necessitates changing the print pattern on fabrics, a series of adaptation processes including image acquisition, drawing to generate a design draft, model training, and model verification. This entire process typically takes several hours. Current technology requires the inkjet printing equipment used for digital printing to perform this entire process, wasting the normal operating time of the inkjet printing equipment and resulting in excessively high operating costs.

[0004] Therefore, there is an urgent need to provide an AI model training and testing device and inkjet printing equipment for digital printing that can avoid wasting normal operating time and significantly reduce operating costs for inkjet printing equipment. Utility Model Content

[0005] This utility model addresses the shortcomings of existing technologies, such as wasting normal operating time of inkjet printing equipment and resulting in excessively high operating costs. To achieve one objective of this utility model, it provides an AI model training device for acquiring patterns from a printing substrate, comprising:

[0006] A camera, used to capture images of the substrate in the area to be printed;

[0007] A light source is used to provide illumination for the camera to capture images of the substrate in the area to be printed.

[0008] The model training unit is connected to the camera and receives image data of the substrate image of the area to be printed sent by the camera. The model training unit trains the image data to obtain model data that matches the area to be printed.

[0009] Preferably, the AI ​​model training device further includes a controller, which includes a display module and a data input module electrically connected to each other. The data input module is used to receive user input information, and the display module displays the input information.

[0010] Preferably, the display module is a touch screen display, which can display an image of the substrate in the area to be printed, and the data input module includes a data acquisition application and a model testing application, which are displayed on the touch screen display.

[0011] Preferably, the model training unit includes a cloud server and a local server. The cloud server includes a cloud data annotation and AI model training module that completes the annotation of data and training of the model in the cloud. The local server completes the annotation of data and training of the model locally, including a local data annotation and AI model training module. The camera sends image data to the cloud server or the local server, and the camera downloads the model data trained by one of the cloud data annotation and AI model training modules and the local data annotation and AI model training module.

[0012] Preferably, the AI ​​model training device further includes a built-in model inference module disposed in the camera or an external model inference module disposed for executing the inkjet printing device. The built-in model inference module completes image acquisition of the area to be printed and model inference of the model data, and sends the model inference result to the inkjet printing device and the touch screen, and the touch screen displays the model inference result. The external model inference module is the operating system of the inkjet printing device. The external model inference module receives the original grayscale image or model inference result sent by the camera, and performs accelerated model inference on the original grayscale image or model inference result through a model accelerator, and sends the accelerated model inference result to the inkjet printing device and the touch screen, and the touch screen displays the accelerated model inference result.

[0013] Preferably, the built-in model inference module and the external model inference module respectively judge the results of the model inference and the model acceleration inference. If the result is qualified, the model data is qualified and is used by the inkjet printing device to perform digital printing on the area to be printed. If the result is unqualified, the model data is unqualified, and the AI ​​model training device performs iterative work until the model data is qualified.

[0014] Preferably, the camera is equipped with an Ethernet interface and connected to an Ethernet switch via Ethernet. The Ethernet switch is connected to the cloud server and the local server respectively. The light source includes multiple light-emitting light sources, and the camera controls the switching of at least some of the multiple light-emitting light sources.

[0015] Preferably, the AI ​​model training device further includes a driving component, the driving component including a driver, wherein the camera controls the driver to control the transfer of the substrate.

[0016] Preferably, the printing substrate is fabric, the camera is an AI camera, the AI ​​camera is equipped with a GPIO, UART, or I2C interface for connecting to and controlling the driver and the light source respectively, and the AI ​​camera is equipped with a Type-C interface for connecting to the display module.

[0017] To achieve another objective of this utility model, an inkjet printing device is provided, comprising: a printing carriage and a printhead disposed on the printing carriage, wherein the inkjet printing device further comprises the AI ​​model training device described in any of the preceding claims.

[0018] The beneficial effects of this utility model are as follows:

[0019] The AI ​​model training device and inkjet printing equipment for collecting substrate patterns provided by this utility model can independently complete the changeover adaptation process required for substrate pattern changeover without occupying the inkjet printing equipment for printing substrate patterns to perform the entire process. This not only saves the inkjet printing equipment the time required for changeover, which helps to significantly reduce its operating costs, but also allows for sufficient testing and verification before changeover, thereby improving production quality. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of this utility model, the drawings used in the embodiments of this utility model will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, and these are all within the protection scope of this utility model.

[0021] Figure 1 A schematic diagram of the structure of the AI ​​model training device for collecting patterns on a printing substrate provided in an embodiment of this utility model;

[0022] Figure 2 The diagram shows the network architecture of the AI ​​model training device for collecting patterns from printed substrates, as described in this utility model.

[0023] Figure 3A schematic diagram of the construction of the AI ​​model training device for collecting patterns on a printing substrate provided in this embodiment of the utility model, and the data flow for implementing model inference;

[0024] Figure 4 The schematic diagram of the construction of the AI ​​model training device for collecting patterns of printed materials provided in this embodiment of the utility model and the data flow of the implementation of accelerated model inference is shown below.

[0025] Explanation of reference numerals in the attached figures:

[0026] 1- Camera; 2- Unwinding / rewinding device; 21- Unwinding roller; 22- Rewinding roller; 23- Drive assembly; 231- Driver; 232- Drive roller; 3- Light source; 4- Model training unit; 41- Cloud server; 42- Local server; 5- Controller; 51- Display module; 52- Data receiving module; 521- Data acquisition application; 522- Model testing application; 6- Ethernet switch; 7- Substrate; 100- Fixed column; 101- Lifting mechanism; 110- Substrate platform; 120- Substrate; 102- Support frame; 130- Light source; 131- Knob; 140- Communication unit; 141- Antenna; 150- Camera; 160- Power cord; 170- Communication line. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the embodiments of this utility model clearer, the technical solutions of the embodiments of this utility model will be clearly and completely described below with reference to the accompanying drawings. It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. In the description of this utility model, it should be understood that the terms "center," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this utility model. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element limited by the phrase "comprising..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. Unless otherwise specified, embodiments of the present invention and the various features thereof can be combined with each other, all within the protection scope of the present invention.

[0028] This utility model provides an AI model training device for acquiring patterns from a printing substrate, comprising:

[0029] A camera, used to capture images of the substrate in the area to be printed;

[0030] A light source is used to provide illumination for the camera to capture images of the substrate in the area to be printed.

[0031] The model training unit is connected to the camera and receives image data of the substrate image of the area to be printed sent by the camera. The model training unit trains on the image data to obtain model data that matches the area to be printed. It should be noted that the model training unit here is a hardware device that can use existing image model training techniques to achieve image matching. This invention mainly involves installing a novel AI model training device for acquiring substrate patterns on existing printing equipment. This device can significantly reduce the changeover time required for substrate pattern changes.

[0032] Preferably, the AI ​​model training device further includes a controller, which comprises an electrically connected display module and a data input module. The data input module receives user input information, and the display module displays the input information. These are all common hardware module connections, and their display and input methods are also common in electronic devices.

[0033] Please see Figure 2 In the structural diagram of the AI ​​model training device of this utility model, the AI ​​model training device includes: a fixed column 100 provided on one side of the substrate platform 110 for supporting the substrate 120; a lifting mechanism 101 inserted into the fixed column 100; a fixed support frame 102 connected to one end of the lifting mechanism 101; and a slidable AI model training component nested on the support frame 102. The AI ​​model training component, from top to bottom, includes: a knob 131 for limiting or moving a camera 150; a communication unit 140 connected to the camera 150; multiple antennas 141 protruding on one side of the 140; and a light source 130 arranged side-by-side with the communication unit 140 for providing illumination for the camera 150 during operation. The communication unit 140 establishes a communication connection with a host computer (not shown) via a communication line 170. The entire AI model training component is connected to a power source via a power cord 160 to provide operating power.

[0034] Preferably, the display module is a touch screen display, which can display an image of the substrate in the area to be printed, and the data input module includes a data acquisition application and a model testing application, which are displayed on the touch screen display.

[0035] Preferably, the model training unit includes a cloud server and a local server. The cloud server includes a cloud data annotation and AI model training module that completes the annotation of data and training of the model in the cloud. The local server completes the annotation of data and training of the model locally, including a local data annotation and AI model training module. The camera sends image data to the cloud server or the local server, and the camera downloads the model data trained by one of the cloud data annotation and AI model training modules and the local data annotation and AI model training module.

[0036] Preferably, the AI ​​model training device further includes a built-in model inference module disposed in the camera or an external model inference module disposed for executing the inkjet printing device. The built-in model inference module completes image acquisition of the area to be printed and model inference of the model data, and sends the model inference result to the inkjet printing device and the touch screen, and the touch screen displays the model inference result. The external model inference module is the operating system of the inkjet printing device. The external model inference module receives the original grayscale image or model inference result sent by the camera, and performs accelerated model inference on the original grayscale image or model inference result through a model accelerator, and sends the accelerated model inference result to the inkjet printing device and the touch screen, and the touch screen displays the accelerated model inference result.

[0037] Preferably, the built-in model inference module and the external model inference module respectively judge the results of the model inference and the model acceleration inference. If the result is qualified, the model data is qualified and is used by the inkjet printing device to perform digital printing on the area to be printed. If the result is unqualified, the model data is unqualified, and the AI ​​model training device performs iterative work until the model data is qualified.

[0038] Preferably, the camera is equipped with an Ethernet interface and connected to an Ethernet switch via Ethernet. The Ethernet switch is connected to the cloud server and the local server respectively. The light source includes multiple light-emitting light sources, and the camera controls the switching of at least some of the multiple light-emitting light sources.

[0039] Preferably, the AI ​​model training device further includes a driving component, the driving component including a driver, wherein the camera controls the driver to control the transfer of the substrate.

[0040] Preferably, the printing substrate is fabric, the camera is an AI camera, the AI ​​camera is equipped with a GPIO, UART, or I2C interface for connecting to and controlling the driver and the light source respectively, and the AI ​​camera is equipped with a Type-C interface for connecting to the display module.

[0041] The following detailed description uses digital printing as an example of printing substrates; other common flexible printing substrates are also within the scope of protection of this utility model. (Reference) Figure 1The AI ​​model training device for acquiring patterns on the substrate includes a camera 1, a light source 3, and a model training unit 4. The camera 1 captures images of the area of ​​the substrate 7 to be printed in real time to obtain image data; this image data can be contained in the captured image file. The unwinding / winding device 2 of the inkjet printer is connected to the camera 1 and, under the control of the camera 1, winds up and unwinds the substrate 7. The light source 3 is connected to the camera 1 and, under the control of the camera 1, provides illumination to at least the area of ​​the substrate 7 to be printed. Preferably, the unwinding / winding device 2 and the light source 3 are connected to the camera 1 via an embedded system peripheral interface or a hardware communication interface (described further below) and via cables. The model training unit 4 is connected to the camera 1 and receives the image data. The model training unit 4 trains on the image data to obtain model data matching the area to be printed. Therefore, the AI ​​model training device for collecting patterns on the substrate provided by this utility model can independently complete the changeover adaptation process required for fabric printing, without occupying the inkjet printing equipment that performs digital printing operations to execute the entire process. This not only saves the time required for inkjet printing equipment to perform changeovers, which helps to significantly reduce its operating costs, but also allows for sufficient testing and verification before changeovers, thereby improving production quality.

[0042] Please refer to Figure 1 Preferably, the AI ​​model training device for acquiring the pattern of the printed substrate further includes a controller 5. The controller 5 includes a display module 51 and a data receiving module 52 electrically connected. The data receiving module 52 is used to receive input information from the data input module, and the display module 51 displays the input information. In this way, the display module 51 can be used to intuitively observe the real-time images captured by the camera 1, deduce the model process, and facilitate the user to input information such as parameters related to model adaptation.

[0043] Please refer to Figure 1 Specifically, the display module 51 is a touch screen that displays the image corresponding to the image data in real time. The data input module includes a data acquisition application 521 and a model testing application 522. The touch screen displays the input information entered through the corresponding interface of each application. In this way, users can not only conveniently input information by touch, but also intuitively input relevant parameters for data acquisition and model testing of the image data through the interface of the corresponding application.

[0044] Please refer to Figure 1Specifically, the model training unit 4 includes a cloud server 41 and a local server 42. The cloud server 41 includes a cloud data annotation and AI model training module that completes the annotation of data and model training in the cloud. The local server 42 includes a local data annotation and AI model training module that completes the annotation of data and model training locally. The camera 1 sends image data to the cloud server 41 or the local server 42, and the camera 1 downloads model data trained by either the cloud data annotation and AI model training module or the local data annotation and AI model training module. Therefore, the local or cloud training of the model can be flexibly selected based on factors such as the speed of model training, the ease of uploading image files and downloading trained model files, and the complexity of the training model.

[0045] Please refer to the reference. Figure 3 and Figure 4 Preferably, the AI ​​model training and testing system further includes a built-in model inference module installed in the camera 1 or an external model inference module installed in the inkjet printing device used for digital printing. The built-in model inference module completes image acquisition of the area to be printed and model inference of the model data, and sends the model inference results to the inkjet printing device and the touch screen. The touch screen displays the model inference results. The external model inference module is the operating system of the inkjet printing device. In this invention, the operating system is specifically a Windows system. The external model inference module receives the original grayscale image or model inference results sent by the camera 1, and performs accelerated model inference on the original grayscale image or model inference results through a model accelerator. In this invention, the model accelerator specifically uses a GPU. In addition, different model accelerators such as ASIC, TPU, and NPU can be selected according to different operating systems. The accelerated model inference results are sent to the inkjet printing device and the touch screen, and the touch screen displays the accelerated model inference results. More specifically, the environment for running the inference model is divided into a Windows environment and an environment such as an AI camera environment. Figure 3 As shown, for scenarios with low computational requirements, such as when the substrate 7 is jacquard fabric, the model inference module is deployed in the AI ​​camera. The AI ​​camera completes the entire process of image acquisition, model inference, and result transmission. Figure 4 As shown, for scenarios with high computational requirements, such as when the substrate 7 is knitted fabric, the AI ​​camera sends the original grayscale image or binary image (model inference result) to the Windows system, and the model inference is accelerated by the GPU of the Windows system; thus, the purpose of flexibly using different model inference modules to meet the model inference requirements according to the different computational requirements of different substrates 7 is achieved.

[0046] Please refer to Figure 1Preferably, the built-in model inference module and the external model inference module respectively judge the results of model inference and model acceleration inference. If the result is qualified, the model data is qualified and is used by the inkjet printer to perform digital printing on the area to be printed. If the result is unqualified, the model data is unqualified, and the AI ​​model training and testing system performs iterative work until the model data is qualified. In this way, qualified model data can be trained in a highly automated iterative manner, thereby ensuring that qualified model data is transmitted to the inkjet printer's printing system and processed by RIP for the nozzles of the printing system to perform accurate printing.

[0047] Please refer to Figure 1 Preferably, camera 1 is equipped with an Ethernet interface and connected to Ethernet switch 6 via Ethernet. Ethernet switch 6 is connected to cloud server 41 and local server 42 respectively. Therefore, it can ensure that camera 1 reliably sends image data to cloud server 41 or local server 42 and downloads trained model files containing model data from cloud server 41 or local server 42. Light source 3 includes multiple light sources. Camera 1 controls the switching of at least some of the multiple light sources, that is, camera 1 controls the switching of some or all of the light sources in the light source combination, so that camera 1 can use different wavelengths and light spots to emit light to the area to be printed according to different substrates 7, thereby obtaining high-quality image data.

[0048] Please refer to the reference. Figure 1 Specifically, the unwinding / rewinding device 2 includes an unwinding roller 21, a take-up roller 22, and at least one drive assembly 23 located between the unwinding roller 21 and the take-up roller 22 in the direction of movement of the substrate 7. The drive assembly 23 includes a driver 231, a motor, and a pair of drive rollers 232, which are specifically drive rollers. The camera 1 controls the driver 231 to drive the motor to rotate, thereby moving the substrate 7 located between the pair of drive rollers 232, thus realizing the unwinding and rewinding of the substrate 7. In this way, the camera 1 can automatically adjust the moving speed of the drive assembly 23 to drive the substrate 7 to address the irregular elastic deformation caused by the difference in elastic coefficients of different substrates 7, thereby meeting its actual need to capture high-quality images and obtain high-quality image data.

[0049] Please refer to Figure 1Specifically, the printing substrate 7 is fabric, including jacquard fabric, knitted fabric, lace fabric, etc. Camera 1 is an AI camera. This means that an AI camera uses artificial intelligence (AI) technology applied to model training and testing, i.e., inference, as described in this specification. For example, the AI ​​camera in this invention uses a host unit equipped with three cameras, with a maximum field of view of 1.8 meters, a physical resolution of 0.1 mm, a resolution of 8000*2000 for each camera, a total resolution of 24000*2000, an imaging frame rate of 5 FPS, a model inference and transmission frame rate of 1 FPS, and a model segmentation accuracy of 0.4 mm. The AI ​​camera is equipped with GPIO, UART, or I2C interfaces for connecting to and controlling the driver 231 and the light source 3, respectively. The AI ​​camera is also equipped with a Type-C interface for connecting to the display module. Therefore, it can adapt to various types of fabric, and the AI ​​camera can reliably connect to the driver 231, the light source 3, and the display module with low-latency communication.

[0050] The functions of this AI camera are: connecting to a touchscreen display to show the App interface and respond to user input; taking photos and displaying real-time images; developing AI models and displaying the results; controlling the switching of some or all of the light sources in a light source combination; controlling the motor driver to rotate the motor, enabling the unwinding and rewinding of fabric; sending image data to a cloud server or local server; and downloading trained model files from the cloud server or local server.

[0051] Iterative workflow:

[0052] ① AI camera acquires images -> ② Sends images to local server / cloud server -> ③ Labels data -> ④ Trains model -> ⑤ AI camera downloads model -> ⑥ AI camera acquires real-time images, infers model, displays inference results -> Identifies problems -> ①.

[0053] Another objective of this invention is to provide an inkjet printing device, including a printing carriage and a printhead mounted on the printing carriage. The printing carriage can be adapted to Onepass or multipass printing as needed. The inkjet printing device also includes an AI model training device for acquiring patterns from the substrate, as described above. The printhead ejects ink from the area to be printed based on qualified model data. It is understood that the inkjet printing device operates relatively independently from the AI ​​model training device for acquiring patterns from the substrate and can communicate via an RJ45 interface network cable. The inkjet printing device can obtain data from any... The beneficial effects of an AI model training device for collecting substrate patterns in digital printing are as follows: the printing system and operating system spray ink to complete the printing process based on the qualified model data obtained by the AI ​​model training device through training and deduction. Before the AI ​​model training device for collecting substrate patterns completes the training and deduction of new models, the printing system and operating system can perform the printing tasks of known trained and deduced models in parallel, without wasting the production time of the inkjet printing equipment. Furthermore, sufficient testing and verification can be carried out before model changeover, thus improving the overall production quality of the inkjet printing equipment.

[0054] Finally, it should be noted that the above description is only used to illustrate the technical solution of this utility model, and not to limit it. Although this utility model has been described in detail, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this utility model. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this utility model should be included within the protection scope of this utility model.

Claims

1. An AI model training device for acquiring patterns from a printing substrate, characterized in that, include: A camera, used to capture images of the substrate in the area to be printed; A light source is used to provide illumination for the camera to capture images of the substrate in the area to be printed. The model training unit is connected to the camera and receives image data of the substrate image of the area to be printed sent by the camera. The model training unit trains the image data to obtain model data that matches the area to be printed.

2. The AI ​​model training device according to claim 1, characterized in that, The AI ​​model training device also includes a controller, which includes a display module and a data input module electrically connected to each other. The data input module is used to receive user input information, and the display module displays the input information.

3. The AI ​​model training device according to claim 2, characterized in that, The display module is a touch screen, which can display an image of the substrate in the area to be printed. The data input module includes a data acquisition application and a model testing application, which are displayed on the touch screen.

4. The AI ​​model training device according to claim 3, characterized in that, The model training unit includes a cloud server and a local server. The cloud server includes a cloud data annotation and AI model training module that completes the annotation of data and training of the model in the cloud. The local server completes the annotation of data and training of the model locally, including a local data annotation and AI model training module. The camera sends image data to the cloud server or the local server, and the camera downloads the model data trained by one of the cloud data annotation and AI model training modules and the local data annotation and AI model training module.

5. The AI ​​model training device according to claim 4, characterized in that, The AI ​​model training device further includes a built-in model inference module disposed in the camera or an external model inference module disposed for executing the inkjet printing device. The built-in model inference module completes image acquisition of the area to be printed and model inference of the model data, and sends the model inference result to the inkjet printing device and the touch screen. The touch screen displays the model inference result. The external model inference module is the operating system of the inkjet printing device. The external model inference module receives the original grayscale image or model inference result sent by the camera, and performs accelerated model inference on the original grayscale image or model inference result through a model accelerator, and sends the accelerated model inference result to the inkjet printing device and the touch screen. The touch screen displays the accelerated model inference result.

6. The AI ​​model training device according to claim 5, characterized in that, The built-in model inference module and the external model inference module respectively judge the results of the model inference and the model acceleration inference. If the result is qualified, the model data is qualified and is used by the inkjet printing device to perform digital printing on the area to be printed. If the result is unqualified, the model data is unqualified, and the AI ​​model training device performs iterative work until the model data is qualified.

7. The AI ​​model training device according to claim 4, characterized in that, The camera is equipped with an Ethernet interface and connected to an Ethernet switch via Ethernet. The Ethernet switch is connected to the cloud server and the local server respectively. The light source includes multiple light sources, and the camera controls the switching of at least some of the multiple light sources.

8. The AI ​​model training device according to claim 2, characterized in that, The AI ​​model training device also includes a driving component, which includes a driver, and the camera controls the driver to control the transport of the substrate.

9. The AI ​​model training device according to claim 8, characterized in that, The printing substrate is fabric, the camera is an AI camera, the AI ​​camera is equipped with GPIO, UART or I2C interfaces for connecting and controlling the driver and the light source respectively, and the AI ​​camera is equipped with a Type-C interface for connecting to the display module.

10. An inkjet printing apparatus, comprising a printing carriage and a printhead disposed on the printing carriage, characterized in that, The inkjet printing equipment also includes the AI ​​model training device as described in any one of claims 1 to 9.

Images