Customized mural generation system based on image AI decomposition and assembly equipment thereof

Abstract

The invention discloses a customized mural generation system based on image AI decomposition and assembly equipment thereof, and relates to the technical field of graphic processing. An element extraction module automatically identifies and extracts various elements in an image to split the complex image into a plurality of sub-images or image areas with independent semanteme; and the mosaicking processing module performs mosaicking processing on the split sub-images or image areas by using an image processing algorithm, wherein the mosaicking processing comprises generation of multi-order gray scales and multi-order colors. A horizontal multi-joint robot and a face recognition image processing technology are combined, an image in a shot or uploaded picture is intercepted through an artificial intelligence image processing technology, and the intercepted image is subjected to mosaic processing by using an accurate image processing algorithm; and then a controller is used for controlling the two parallel horizontal multi-joint robots to cooperate to quickly grab pixel pieces according to the information of the mosaic puzzle so as to splice an image on a painting plate.

Description

Technical Field

[0001] The present invention relates to the field of graphics processing technology, and in particular to a customized mural generation system based on image AI decomposition and an assembly device thereof. Background Art

[0002] The customized mural generation system based on image AI decomposition uses advanced algorithms and models to deeply analyze the images or design concepts provided by users, and generates customized murals based on the analysis results. However, the existing mural generation system directly prints out customized murals, but the generated murals have a short preservation time and have certain defects.

[0003] The defects of existing mural generation systems and assembly equipment are: 1. In the patent document CN117291479B, the device portrait architecture of the entire UPFC is mainly constructed to clearly and comprehensively display the multi-dimensional operating characteristics of the device, but it does not take into account that the existing mural generation system directly prints customized murals, and the murals have a short preservation time; 2. In the patent document CN101159064B, the main consideration is how to automatically generate a portrait with personalized features based on the real image provided by the user, while reducing the amount of calculation and improving the efficiency of portrait generation. It does not take into account that the existing mural generation system cannot switch different pixel slices for assembly according to different customized styles, and has poor flexibility; 3. In the patent document CN112434191B, the main consideration is how to improve the accuracy of the portraits of small and micro enterprises, but it does not take into account that the stability of the image assembly of the existing mural generation system cannot be guaranteed. Summary of the invention

[0004] The purpose of the present invention is to provide a customized mural generation system based on image AI decomposition and its assembly equipment to solve the problems raised in the above background technology.

[0005] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: a customized mural generation system based on image AI decomposition, comprising an image acquisition unit, an image recognition unit, an image decomposition unit, an image customization unit, an image reorganization unit, an image display unit and a control unit, wherein the image acquisition unit acquires the image to be generated by shooting with a camera or uploading an existing picture, the image recognition unit intercepts the image in the picture by face recognition image processing technology, the image decomposition unit decomposes the intercepted image based on a deep learning algorithm, the image customization unit customizes the decomposed image according to the needs and preferences of the user by using the generative adversarial network technology, the image reorganization unit reorganizes the customized image according to certain rules and layout by using the image fusion technology to generate the final customized mural image, the image display unit displays the final customized mural image and the process of assembling the image through the display screen, and the control unit controls the horizontal multi-joint robot to clamp the pixel slices and piece together the image on the painting plate through the controller; The image decomposition unit includes an element extraction module, a mosaic processing module and a feature analysis module. The element extraction module automatically identifies and extracts various elements in the image to split the complex image into multiple sub-images or image areas with independent semantics. The mosaic processing module uses image processing algorithms to mosaic the split sub-images or image areas. The feature analysis module uses clustering algorithms, edge detection algorithms, machine learning and deep learning technologies to perform feature analysis on each sub-image or image area. Feature distinction includes color, texture, shape and edge. The mosaic process includes generating multi-level grayscale and multi-level color.

[0006] Preferably, the image customization unit includes a style conversion module and a color adjustment module. The style conversion module uses generative adversarial network technology to convert the extracted image elements into an artistic style specified by a user, and the artistic styles include oil painting style, watercolor style and cartoon style. The color adjustment module uses image processing technology and screen post-processing technology to change the brightness, contrast and saturation of the color of the image elements.

[0007] Preferably, the image reconstruction unit includes a layout design module and a fusion optimization module. The layout design module provides a variety of preset layout templates for users to choose from through automatic layout or automatically adjusts the position, size and proportional relationship of image elements according to user-specified layout requirements. The fusion optimization module uses image fusion technology to further optimize the overall image. The optimization processing includes adjusting light and shadow effects and enhancing image layering.

[0008] Preferably, a customized mural assembly device based on image AI decomposition includes an assembly table, a transparent baffle and a storage plate are arranged on the top of the assembly table, the storage plate is located behind the transparent baffle, the transparent baffle and the storage plate are combined into a circular structure, a camera is installed on the front of the transparent baffle, a display screen is installed on the front of the storage plate, two parallel horizontal multi-joint robots are installed on the top of the assembly table, a plurality of groups of visual sensors are arranged on the surface of the horizontal multi-joint robots, and the visual sensors, cameras and display screens are electrically connected.

[0009] Preferably, a support plate is installed on the inner wall of the assembly table, and a symmetrically arranged servo motor is installed on the back of the support plate, a storage barrel is installed on the output end of the servo motor, and a connecting block is installed on the inner wall of the storage barrel. The connecting block has a square structure, and an electric push rod is installed on the outer surface of the connecting block. The output end of the electric push rod passes through the outer surface of the storage barrel and is installed with a connecting plate, and the outer surfaces of the four groups of connecting plates are installed with storage frames, and the inner surfaces of the storage frames are embedded with evenly arranged electromagnets, and evenly arranged pixel pieces are adsorbed on the electromagnets, and the pixel pieces in different storage frames have different colors. The top of the assembly table is penetrated by symmetrically arranged storage slots, which are embedded with one of the groups of storage frames, and the storage slots are located in front of the horizontal multi-joint robot.

[0010] Preferably, the top wall of the assembly table is provided with an adjustment component, and symmetrically arranged assembly plates are installed through the top of the assembly table. Drawing trays are slidably placed on the top of two groups of assembly plates, and one side outer wall of one group of drawing trays is provided with evenly arranged sockets, and one side outer wall of the other group of drawing trays is provided with evenly arranged insertion rods, and the insertion rods are adapted to the sockets.

[0011] Preferably, the bottom wall of the drawing tray is provided with evenly arranged limiting blocks, and one side outer surface of the pixel sheet is provided with a limiting hole, and the limiting hole is matched with the limiting block.

[0012] Preferably, the adjustment component includes three groups of supporting plates installed on the top wall of the assembly table, wherein two groups of supporting plates are installed with variable frequency motors on one side outer walls, reciprocating screws are installed on the output ends of the variable frequency motors, and one ends of the two groups of reciprocating screws are respectively connected to the outer walls of both sides of another group of supporting plates through shafts, the outer surfaces of the reciprocating screws are sleeved with sliding sleeves, the outer surfaces of the sliding sleeves are installed with connecting rods, a slide groove is provided through the top of the assembly table, and the slide groove is located in the middle of the two groups of horizontal multi-joint robots, and the top of the connecting rod passes through the inside of the slide groove and is connected to the bottom of the assembly plate.

[0013] Preferably, the color levels of the pixel slices in different placement frames are different.

[0014] Preferably, the transparent shield is of an openable and closable design, and the assembly plate restricts the drawing tray in the front and rear directions.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention combines a horizontal multi-joint robot with face recognition image processing technology, uses artificial intelligence image processing technology to capture images in photographed or uploaded pictures, uses precise image processing algorithms to mosaic the captured images, and then uses a controller to control two parallel horizontal multi-joint robots to cooperate and quickly grab pixel pieces to piece together an image on a drawing board according to the information of the mosaic puzzle.

[0016] 2. The present invention places pixel pieces of different colors in storage frames at different positions according to different styles of customized pictures by users. When assembling the image, the servo motor is started to drive the storage barrel to rotate, and the position of the storage frame with pixel pieces of different colors is adjusted. Then, the electric push rod is started to drive the storage frame to engage with the storage slot, and then the electromagnet in the storage frame is powered off, so that the horizontal multi-joint robot can grab the pixel pieces in the storage frame to piece together the customized image on the drawing board, which can improve the flexibility of use and the diversity of customization to a certain extent.

[0017] 3. The present invention fixes the pixel sheet in the storage frame by means of magnetic attraction, thereby facilitating the fixing of pixel sheets of different colors in the storage frame at different positions, and at the same time improving the stability of subsequent image assembly by means of the setting of limiting holes and limiting blocks.

[0018] 4. The present invention drives the assembly plate to move by adjusting the setting of the components, thereby adjusting the position of the drawing plate, and thus to a certain extent can improve the convenience of image assembly. At the same time, the assembled drawing plates are combined together by setting the plug rods and the sockets, thereby improving the speed and convenience of robot assembly. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is a system diagram of the present invention; Figure 2 A diagram showing the composition of the image customization unit of the present invention; Figure 3 is a diagram showing the structure of an image reorganization unit of the present invention; Figure 4 It is a schematic diagram of the overall structure of the present invention; Figure 5 It is a schematic diagram of the assembly structure of the assembly platform of the present invention; Figure 6 It is a schematic diagram of the assembly structure of the assembly platform and the horizontal multi-joint robot of the present invention; Figure 7 It is a schematic diagram of the planar assembly structure of the storage tube and the storage frame of the present invention; Figure 8 It is a schematic diagram of the planar assembly structure of the adjustment component of the present invention; Fig. 9It is a schematic diagram of the planar assembly structure of the electric push rod, the storage frame and the pixel sheet of the present invention; Fig.10 It is a schematic diagram of the planar assembly structure of the storage frame and the storage slot of the present invention.

[0020] In the figure: 1. assembly table; 2. transparent baffle; 3. camera; 4. storage plate; 5. display screen; 6. horizontal multi-joint robot; 7. support plate; 8. servo motor; 9. storage tube; 10. storage slot; 11. connecting block; 12. electric push rod; 13. connecting plate; 14. storage frame; 15. electromagnet; 16. pixel sheet; 17. limit hole; 18. assembly plate; 19. drawing plate; 20. socket; 21. plug rod; 22. limit block; 23. slide groove; 24. load-bearing plate; 25. frequency conversion motor; 26. reciprocating screw rod; 27. sleeve; 28. connecting rod. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] In the description of the present invention, it should be noted that the terms "upper", "lower", "inner", "outer", "front end", "rear end", "two ends", "one end", "the other end" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance.

[0023] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "provided with", "connected", etc. should be understood in a broad sense. For example, "connected" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0024] See also Figure 1 , Figure 4 and Figure 5, an embodiment provided by the present invention: a customized mural generation system based on image AI decomposition, comprising an image acquisition unit, an image recognition unit, an image decomposition unit, an image customization unit, an image reorganization unit, an image display unit and a control unit, wherein the image acquisition unit acquires an image to be generated by shooting or uploading an existing image through a camera 3, the image recognition unit intercepts an image in the image through a face recognition image processing technology, the image decomposition unit decomposes the intercepted image based on a deep learning algorithm, the image customization unit customizes the decomposed image according to the needs and preferences of the user using a generative adversarial network technology, the image reorganization unit reorganizes the customized image according to certain rules and layout using an image fusion technology to generate a final customized mural image, and the image display unit generates a customized mural image by a control unit. The display screen 5 displays the final customized mural image and the process of assembling the image. The control unit controls the horizontal multi-joint robot 6 through the controller to clamp the pixel slices 16 to piece together the image on the drawing board 19. The image decomposition unit includes an element extraction module, a mosaic processing module and a feature analysis module. The element extraction module automatically identifies and extracts various elements in the image to split the complex image into multiple sub-images or image areas with independent semantics. The mosaic processing module uses an image processing algorithm to mosaic the split sub-images or image areas. The feature analysis module performs feature analysis on each sub-image or image area through clustering algorithms, edge detection algorithms, machine learning and deep learning technologies. Feature distinction includes color, texture, shape and edge. Mosaic processing includes generating multi-level grayscale and multi-level color.

[0025] Furthermore, the original image material is obtained through the camera 3 or uploaded by a mobile phone, and then the collected image is preprocessed by using AI technology, including operations such as removing noise, adjusting color balance, and enhancing contrast, so as to improve the quality and clarity of the image and provide a good basis for subsequent decomposition and processing. If it is a face image, the face recognition image processing technology is used to capture the image in the picture, and then the convolutional neural network and recurrent neural network technologies in deep learning are used to decompose the preprocessed image, automatically identify and extract various elements in the image, such as people, objects and background, and split the complex image into multiple sub-images or image areas with independent semantics. For example, in a landscape mural, elements such as the sky, mountains, trees and rivers can be accurately separated, and then each sub-image or image area is subjected to feature analysis to extract its color, texture, shape and edge, etc. Feature information, and mosaic processing is performed on the selected sub-image or image area, so as to facilitate the subsequent horizontal multi-joint robot 6 to grab the pixel slices 16 to piece together a customized image on the painting plate 19. In the process of image processing, the generative adversarial network technology can be used to convert the extracted image elements into the artistic style specified by the user, such as oil painting style, watercolor style or cartoon style, etc. The user can select different style templates through simple operations, or input specific style parameters to make the generated image have a unique artistic effect. In the process of image generation, the user is allowed to modify, add or replace the image elements, and adjust the overall color of the mural or the color of each element according to the user's preference for color. Finally, the customized image elements are reorganized according to certain rules and layouts to generate the final customized mural image, so as to achieve highly personalized customization; During the reconstruction process, image fusion technology is used to make the transition between each image element natural and seamless, avoiding obvious splicing marks. At the same time, the overall image is further optimized, such as adjusting the light and shadow effects, enhancing the sense of layering, etc., to improve the artistic quality and realism of the mural image.

[0026] See also Figure 2 An embodiment of the present invention is: a customized mural generation system based on image AI decomposition, the image customization unit includes a style conversion module and a color adjustment module, the style conversion module uses generative adversarial network technology to convert the extracted image elements into an artistic style specified by the user, the artistic styles include oil painting style, watercolor style and cartoon style, and the color adjustment module uses image processing technology and screen post-processing technology to change the brightness, contrast and saturation of the image element color.

[0027] Furthermore, the generative adversarial network technology includes a generator and a discriminator. The generator is responsible for generating an image with a corresponding style based on the input image elements and the artistic style requirements specified by the user, and the discriminator is used to determine whether the generated image conforms to the real characteristics of the artistic style. By continuously feeding back to the generator for adjustment, the final generated image can realistically present the artistic style expected by the user, including black and white pictures, optimized style transfer, watercolor style transfer and cartoon style transfer, etc., and use image processing technology and screen post-processing technology to change the relevant parameters of the color of image elements, such as changing the brightness, contrast and saturation of the color by calculating the RGB values ​​of the image pixels, so as to achieve a more ideal visual effect.

[0028] See also Figure 3 An embodiment of the present invention is: a customized mural generation system based on image AI decomposition, the image reconstruction unit includes a layout design module and a fusion optimization module, the layout design module provides a variety of preset layout templates for users to choose through automatic layout or automatically adjusts the position, size and proportion relationship of image elements according to the layout requirements specified by the user, and the fusion optimization module uses image fusion technology to further optimize the overall image, and the optimization processing includes adjusting light and shadow effects and enhancing image layering.

[0029] Furthermore, the preset layout templates include symmetrical layout templates, center-focused layout templates, and scattered layout templates. The symmetrical layout template is characterized in that the image elements are symmetrically distributed left and right or up and down based on the central axis. The center-focused layout template places the core image element in the center of the picture, and other elements are spread out around it, which can effectively highlight the key elements. The image elements of the scattered layout template are distributed in the picture in a scattered and random manner. Users can browse these preset templates and choose a suitable template based on their initial ideas for customizing murals. They can also accurately adjust the image elements according to the layout requirements specified by the user. For example, if a specific element is expected to occupy a larger proportion of the picture, or certain elements are placed in a specific position relationship, the position, size and proportion relationship of the image elements are automatically adjusted through precise calculation and operation. If the user wants to highlight the human elements in the image, he or she can request to enlarge the human elements and place them near the golden section point of the picture. The layout design module will calculate the appropriate magnification and movement coordinates based on the original size and position of the human elements in the image and the overall size of the picture. Image fusion technology is designed to make each image after customized processing and layout adjustment The elements can achieve a natural transition and seamless splicing to avoid obvious splicing marks. Common image fusion technologies include methods based on weighted average, that is, different weights are given to adjacent image elements according to factors such as the importance or relevance of the elements, and then the pixel value of the fusion area is determined by weighted average calculation to achieve a smooth transition. There are also methods based on multi-resolution analysis, which decompose and reconstruct the image at different resolutions and perform fusion processing at different resolution levels to achieve a better fusion effect, such as enhancing the brightness of the highlight area to make it more eye-catching and highlight the important elements or details in the picture. At the same time, the darkness of the shadow area is appropriately lowered to increase the layering of the shadow and make the picture more three-dimensional. The perspective principle can also be used to blur the image elements at different distances to different degrees, making the distant elements look more blurred, thereby highlighting the depth of the picture. The color saturation, contrast and other parameters of the image elements can also be adjusted to make the elements at different levels visually distinguishable, further enhancing the sense of layering. Through these operations, the customized mural image can present a richer three-dimensional visual effect on the two-dimensional picture.

[0030] See also Figure 4An embodiment of the present invention is: a customized mural assembly device based on image AI decomposition, including an assembly table 1, a transparent baffle 2 and a storage board 4 are arranged on the top of the assembly table 1, the storage board 4 is located behind the transparent baffle 2, the transparent baffle 2 and the storage board 4 are combined into a circular structure, a camera 3 is installed on the front of the transparent baffle 2, a display screen 5 is installed on the front of the storage board 4, two parallel horizontal multi-joint robots 6 are installed on the top of the assembly table 1, and a plurality of groups of visual sensors are arranged on the surface of the horizontal multi-joint robot 6, and the visual sensors, the camera 3 and the display screen 5 are electrically connected.

[0031] Furthermore, pictures are taken or uploaded through the camera 3, and the images in the pictures are captured through artificial intelligence image processing technology. The captured images are mosaicked using precise image processing algorithms. If the user wants to generate a black and white picture, the mosaicked picture can be divided into four or five grayscales as needed. Then the controller controls two parallel horizontal multi-joint robots 6 to cooperate in quickly grabbing, transporting and placing four or five grayscale pixel pieces 16 according to the information of the mosaic puzzle, and piece together the image on the drawing tray 19. The picking and placing error of the horizontal multi-joint robot 6 is 0.2㎜.

[0032] See also Figure 6 , Figure 7 , Fig. 9 and Fig.10 An embodiment of the present invention is as follows: a customized mural assembly device based on image AI decomposition, wherein a support plate 7 is installed on the inner wall of an assembly table 1, a symmetrically arranged servo motor 8 is installed on the back of the support plate 7, a storage tube 9 is installed on the output end of the servo motor 8, a connecting block 11 is installed on the inner wall of the storage tube 9, and the connecting block 11 is in a square structure. An electric push rod 12 is installed on the outer surface of the connecting block 11, and the output end of the electric push rod 12 penetrates through the outer surface of the storage tube 9 and is installed with a connecting plate 13, and the outer surfaces of the four groups of connecting plates 13 are all installed with storage frames 14, and the inner surface of the storage frame 14 is embedded with uniformly arranged electromagnets 15, and uniformly arranged pixel pieces 16 are adsorbed on the electromagnets 15, and the pixel pieces 16 in different storage frames 14 have different colors, and a symmetrically arranged storage slot 10 is penetrated through the top of the assembly table 1, and the storage slot 10 is embedded with one group of storage frames 14, and the storage slot 10 is located in front of the horizontal multi-joint robot 6, and the pixel pieces 16 in different storage frames 14 have different color levels.

[0033] Furthermore, according to the different customized picture styles of the user, the servo motor 8 on the support plate 7 is started to drive the storage tube 9 to rotate, thereby changing the position of the storage frame 14 equipped with pixel pieces 16 of different colors. Considering the stability of the servo motor 8 during use, a supporting plate can be added to the support plate 7 to improve the support of the servo motor 8, thereby improving the stability of the servo motor 8 and the storage tube 9 during rotation. Then, the electric push rod 12 on the connecting block 11 is started to drive the storage frame 14 to engage with the storage slot 10, and then the electromagnet 15 in the storage frame 14 is powered off, so as to facilitate the horizontal multi-joint robot 6 to grab the pixel pieces 16 in the storage frame 14 to piece together a customized image on the drawing board 19, which can improve the flexibility during use and the diversity of customization to a certain extent.

[0034] See also Figure 4 , Figure 5 and Figure 8 An embodiment of the present invention is as follows: a customized mural assembly device based on image AI decomposition, wherein the top wall of the assembly table 1 is provided with an adjustment component, and the adjustment component passes through the top of the assembly table 1 and is installed with symmetrically arranged assembly plates 18, and painting trays 19 are slidably placed on the tops of the two groups of assembly plates 18, one side outer wall of one group of painting trays 19 is provided with evenly arranged plug holes 20, and one side outer wall of the other group of painting trays 19 is provided with evenly arranged plug rods 21, and the plug rods 21 are adapted to the plug holes 20, and the bottom wall of the painting tray 19 is provided with evenly arranged limit blocks 22, and the outer surface of one side of the pixel sheet 16 is provided with limit holes 17, and the limit holes 17 are adapted to the limit blocks 22, and the transparent shield 2 is of an openable and closable design, and the assembly plate 18 limits the painting tray 19 front and back.

[0035] Furthermore, in the process of using the horizontal multi-joint robot 6 to assemble images, a visual sensor is used to determine the position of the pixel piece 16 grabbed by the horizontal multi-joint robot 6 and its own movement trajectory, and the assembly process is uploaded to the display screen 5. When assembling images, the horizontal multi-joint robot 6 grabs the pixel piece 16 placed in the storage frame 14 and places it on the drawing tray 19 on the assembly board 18, so that the limiting hole 17 on the pixel piece 16 can fall on the limiting block 22 at the corresponding position of the drawing tray 19, and then the pixel piece 16 can be fixed on the drawing tray 19 by the engagement of the limiting block 22 and the limiting hole 17. After the image is assembled, the two groups of assembly boards 18 are driven close to each other by adjusting the components, so that the insertion rod 21 can be engaged with the insertion hole 20, and then the image assembled by the two parallel horizontal multi-joint robots 6 can be combined together to form a complete image. Finally, the transparent cover 2 is opened to take out the assembled image.

[0036] See also Figure 8An embodiment of the present invention is: a customized mural assembly device based on image AI decomposition, the adjustment component includes three groups of supporting plates 24 installed on the top wall of the assembly table 1, wherein one side outer wall of two groups of supporting plates 24 are installed with frequency conversion motors 25, and the output ends of the frequency conversion motors 25 are installed with reciprocating screws 26, and one end of the two groups of reciprocating screws 26 are respectively connected to the outer walls of both sides of another group of supporting plates 24 through shafts, the outer surface of the reciprocating screw 26 is sleeved with a sliding sleeve 27, and the outer surface of the sliding sleeve 27 is installed with a connecting rod 28, a slide groove 23 is set through the top of the assembly table 1, and the slide groove 23 is located in the middle of the two groups of horizontal multi-joint robots 6, and the top of the connecting rod 28 passes through the inside of the slide groove 23 and is connected to the bottom of the assembly plate 18.

[0037] Furthermore, when using the adjustment component to change the position of the assembly plate 18, first start the variable frequency motor 25 on the carrier plate 24, and then drive the reciprocating screw 26 to rotate, so that the sleeve 27 can drive the connecting rod 28 to move in the slide groove 23, and then the assembly plate 18 can be driven by the connecting rod 28 to move horizontally on the assembly table 1, thereby adjusting the position of the drawing plate 19.

[0038] Working principle: First, use the camera 3 to take a picture or use a mobile phone to upload a picture, then use the artificial intelligence image processing technology to capture the image in the picture, and perform multiple processing on the captured image, and then use the precise image processing algorithm to mosaic the captured image, and then start the servo motor 8 on the support plate 7 according to the different styles and colors selected by the user, and then drive the storage tube 9 to rotate, so that the storage frame 14 with the appropriate color pixel piece 16 is rotated to the top, and then start the electric push rod 12 to drive the storage frame 14 to engage with the storage slot 10, and then give the storage The electromagnet 15 in the object frame 14 is powered off, and then the controller controls the two parallel horizontal multi-joint robots 6 to cooperate and quickly clamp the pixel piece 16 according to the information of the mosaic puzzle, and place it at the appropriate position of the painting plate 19, and then the pixel piece 16 is fixed on the painting plate 19 by the engagement of the limit block 22 and the limit hole 17, and after the image assembly is completed, the assembly plate 18 is driven to move by the adjustment component, and the painting plate 19 is spliced ​​together by the engagement of the jack 20 and the insertion rod 21, thereby completing the image assembly, and finally open the transparent cover 2 to take out the assembled portrait.

[0039] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A customized mural generation system based on image AI decomposition, characterized by: The system comprises an image acquisition unit, an image recognition unit, an image decomposition unit, an image customization unit, an image reorganization unit, an image display unit and a control unit. The image acquisition unit acquires an image to be generated by shooting or uploading an existing image through a camera (3). The image recognition unit captures an image in the image through face recognition image processing technology. The image decomposition unit decomposes the captured image based on a deep learning algorithm. The image customization unit uses generative adversarial network technology to customize the decomposed image according to user needs and preferences. The image reorganization unit uses image fusion technology to reorganize the customized image according to certain rules and layout to generate a final customized mural image. The image display unit displays the final customized mural image and the process of assembling the image through a display screen (5). The control unit controls a horizontal multi-joint robot (6) through a controller to clamp pixel slices (16) and assemble the image on a painting plate (19). The image decomposition unit includes an element extraction module, a mosaic processing module and a feature analysis module. The element extraction module automatically identifies and extracts various elements in the image to split the complex image into multiple sub-images or image areas with independent semantics. The mosaic processing module uses image processing algorithms to mosaic the split sub-images or image areas. The feature analysis module uses clustering algorithms, edge detection algorithms, machine learning and deep learning technologies to perform feature analysis on each sub-image or image area. Feature distinction includes color, texture, shape and edge. The mosaic process includes generating multi-level grayscale and multi-level color.

2. The customized mural generation system based on image AI decomposition according to claim 1 is characterized by: The image customization unit includes a style conversion module and a color adjustment module. The style conversion module uses generative adversarial network technology to convert the extracted image elements into an artistic style specified by the user. The artistic styles include oil painting style, watercolor style and cartoon style. The color adjustment module uses image processing technology and screen post-processing technology to change the brightness, contrast and saturation of the color of the image elements.

3. The customized mural generation system based on image AI decomposition according to claim 1 is characterized by: The image reorganization unit includes a layout design module and a fusion optimization module. The layout design module provides a variety of preset layout templates for users to choose through automatic layout or automatically adjusts the position, size and proportion of image elements according to the layout requirements specified by the user. The fusion optimization module uses image fusion technology to further optimize the overall image. The optimization processing includes adjusting light and shadow effects and enhancing image layering.

4. A customized mural assembly device based on image AI decomposition, applicable to a customized mural generation system based on image AI decomposition as claimed in any one of claims 1 to 3, characterized in that: The invention comprises an assembly table (1), wherein a transparent baffle (2) and a storage plate (4) are arranged on the top of the assembly table (1), the storage plate (4) is located behind the transparent baffle (2), the transparent baffle (2) and the storage plate (4) are combined into a circular structure, a camera (3) is installed on the front of the transparent baffle (2), a display screen (5) is installed on the front of the storage plate (4), and two parallel horizontal multi-joint robots (6) are installed on the top of the assembly table (1), a plurality of groups of visual sensors are arranged on the surface of the horizontal multi-joint robots (6), and the visual sensors, the camera (3) and the display screen (5) are electrically connected.

5. The customized mural assembly device based on image AI decomposition according to claim 4 is characterized by: The inner wall of the assembly table (1) is mounted with a support plate (7), the back of the support plate (7) is mounted with symmetrically arranged servo motors (8), the output end of the servo motor (8) is mounted with a storage tube (9), the inner wall of the storage tube (9) is mounted with a connecting block (11), the connecting block (11) is in a square structure, the outer surface of the connecting block (11) is mounted with an electric push rod (12), the output end of the electric push rod (12) passes through the outer surface of the storage tube (9) and is mounted with a connecting plate (13), and the four groups of connecting plates (1 3) are all provided with storage frames (14) on their outer surfaces, and evenly arranged electromagnets (15) are embedded in the inner surfaces of the storage frames (14), and evenly arranged pixel sheets (16) are adsorbed on the electromagnets (15), and the pixel sheets (16) in different storage frames (14) have different colors. The top of the assembly table (1) is provided with symmetrically arranged storage slots (10), and the storage slots (10) are embedded in one group of the storage frames (14). The storage slots (10) are located in front of the horizontal multi-joint robot (6).

6. The customized mural assembly device based on image AI decomposition according to claim 4 is characterized by: The top wall of the assembly table (1) is provided with an adjustment component, and the adjustment component passes through the top of the assembly table (1) and is installed with symmetrically arranged assembly plates (18), and drawing trays (19) are slidably placed on the tops of the two groups of assembly plates (18), one side outer wall of one group of the drawing trays (19) is provided with evenly arranged plug holes (20), and one side outer wall of the other group of the drawing trays (19) is provided with evenly arranged plug rods (21), and the plug rods (21) are adapted to the plug holes (20).

7. The customized mural assembly device based on image AI decomposition according to claim 6 is characterized by: The bottom wall of the drawing plate (19) is provided with evenly arranged limiting blocks (22), and one side outer surface of the pixel sheet (16) is provided with a limiting hole (17), and the limiting hole (17) is adapted to the limiting block (22).

8. The customized mural assembly device based on image AI decomposition according to claim 6 is characterized by: The adjustment component comprises three groups of bearing plates (24) mounted on the top wall of the assembly platform (1), wherein one side outer wall of two groups of bearing plates (24) is mounted with a variable frequency motor (25), the output end of the variable frequency motor (25) is mounted with a reciprocating screw rod (26), and one end of the two groups of reciprocating screw rods (26) are respectively connected to the two side outer walls of another group of bearing plates (24) through shafts, the outer surface of the reciprocating screw rod (26) is sleeved with a sliding sleeve (27), and the outer surface of the sliding sleeve (27) is mounted with a connecting rod (28), a sliding groove (23) is provided through the top of the assembly platform (1), and the sliding groove (23) is located in the middle of the two groups of horizontal multi-joint robots (6), and the top end of the connecting rod (28) passes through the inside of the sliding groove (23) and is connected to the bottom of the assembly plate (18).

9. The customized mural assembly device based on image AI decomposition according to claim 5 is characterized by: The pixel slices (16) in different storage frames (14) have different color levels.

10. The customized mural assembly device based on image AI decomposition according to claim 6, characterized in that: The transparent shield (2) is of an openable and closable design, and the assembly plate (18) limits the front and rear movement of the drawing tray (19).

Images