Hand drawing cabinet three-dimensional model generation method based on AI image recognition

Abstract

The invention discloses a hand-drawn cabinet body three-dimensional model generation method based on AI image recognition, and the method comprises the steps: obtaining a hand-drawn front view image corresponding to a first cabinet body, carrying out the image recognition of a frame line in the hand-drawn front view image, and generating an initial front view; identifying the type labels in the frame lines through the image, and obtaining the types of all frame bodies in the initial front view; according to the type of the frame body, generating the frame body with a corresponding structure in the initial front view, and obtaining a formed front view; performing three-dimensional stretching on the formed front view, and constructing an initial three-dimensional model; generating a drawer three-dimensional model according to the type label corresponding to the drawer frame body; generating a cabinet door three-dimensional model according to the cabinet door type labels in the frame lines and the positions of the cabinet door type labels; and combining the initial three-dimensional model, the drawer three-dimensional model and the cabinet door three-dimensional model, and obtaining a final three-dimensional model according to the size parameters of the first cabinet body. According to the method, the image recognition accuracy can be improved, the cabinet body design sketch can be quickly converted into the three-dimensional model, and the three-dimensional model better meeting the requirements can be constructed.

Description

Technical Field

[0001] The present invention relates to the field of three-dimensional model generation, and in particular to a method for generating a three-dimensional model of a hand-drawn cabinet based on AI image recognition. Background Art

[0002] In the engineering field, 3D models are used for product design, modeling of mechanical parts, etc. They can provide more accurate size and shape information, and can be simulated and analyzed to evaluate the feasibility and performance of the design. Through 3D modeling, the appearance and internal structure of the product can be intuitively displayed, which facilitates communication and collaboration between team members. In recent years, product form design has developed in the direction of intelligence and automation, and has significantly improved the efficiency of conceptual design scheme creation, scheme evaluation accuracy and scheme generation quality, especially in the design of cabinet 3D models. Compared with the use of traditional software to build a 3D model of the cabinet, intelligent 3D model construction can build a 3D model without professionals, saving time and effort, and avoiding users from spending time and energy on special software use training.

[0003] However, the current technology of constructing a cabinet 3D model through intelligent recognition of hand-drawn sketches has errors in the recognition of hand-drawn sketches, which makes the constructed 3D model inaccurate. For example, it can identify the lattice frame of the cabinet, but it cannot accurately distinguish whether it is a storage compartment or a drawer compartment. Summary of the invention

[0004] In view of the above-mentioned defects of the prior art, the technical problem to be solved by the present invention is to provide a method for generating a hand-drawn cabinet three-dimensional model based on AI image recognition, aiming to improve the accuracy of image recognition and construct a three-dimensional model that better meets the requirements.

[0005] To achieve the above object, the present invention provides a method for generating a three-dimensional model of a cabinet by hand-drawing based on AI image recognition, the method comprising:

[0006] Step S1, obtaining a hand-drawn front view image corresponding to a first cabinet, and performing image recognition on frame lines in the hand-drawn front view image; generating an initial front view of the first cabinet according to the recognized frame lines and the thickness of a plate material used to manufacture the first cabinet; wherein the initial front view includes a plurality of frames;

[0007] Step S2: image recognition of the type labels in the frame lines to obtain the types of the frames in the initial front view; according to the types of the frames, a frame of corresponding structure is generated in the initial front view to obtain a formed front view of the first cabinet; wherein the frames include fully enclosed frames, semi-enclosed frames and hollow frames, the semi-enclosed frames are divided into locker frames and drawer frames, the fully enclosed frames are divided into hollow frames and solid frames, and the solid frames include the baseboards of the first cabinet;

[0008] Step S3, performing three-dimensional stretching on the molding front view to construct an initial three-dimensional model of the first cabinet;

[0009] Step S4, generating a corresponding three-dimensional model of the drawer according to the type label corresponding to the drawer frame; generating a corresponding three-dimensional model of the cabinet door according to the cabinet door type label in the frame line and the position of the cabinet door type label; wherein the cabinet door types include sliding door type and swing door type;

[0010] Step S5: combining the initial three-dimensional model, the drawer three-dimensional model and the cabinet door three-dimensional model, and obtaining a final three-dimensional model corresponding to the first cabinet according to the size parameters of the first cabinet.

[0011] Optionally, after step S5, the method further includes:

[0012] According to the structure of the final three-dimensional model, corresponding installation ports and connectors are generated at various parts of the final three-dimensional model; the installation ports and connectors are used to connect the panels at various parts during the actual manufacturing process of the first cabinet to form various frames of the first cabinet.

[0013] Optionally, before step S5, the method further includes:

[0014] Image recognition is used to identify the dimension markings of each frame line in the hand-drawn front view image, so as to obtain the dimension parameters corresponding to the first cabinet.

[0015] Optionally, step S5 includes:

[0016] The initial three-dimensional model, the drawer three-dimensional model and the cabinet door three-dimensional model are combined to generate a default three-dimensional model; a size parameter is input into the default three-dimensional model to adjust the default size, and a final three-dimensional model corresponding to the first cabinet is generated according to the size parameter; wherein the default three-dimensional model has a default size. Optionally, after step S5, the method further includes:

[0017] According to the final three-dimensional model, processing drawings of various panels for manufacturing the first cabinet corresponding to the final three-dimensional model are obtained.

[0018] Optionally, in step S5, combining the initial three-dimensional model and the cabinet door three-dimensional model includes:

[0019] According to the type of the cabinet door three-dimensional model, obtaining a first cabinet door installation component between the cabinet door and the initial three-dimensional model;

[0020] According to the size of the cabinet door, obtaining the number and distribution positions of the first cabinet door installation components;

[0021] The initial three-dimensional model and the cabinet door three-dimensional model are combined according to the first cabinet door installation component, the number of the first cabinet door installation components and the distribution position.

[0022] Optionally, in step S5, combining the drawer three-dimensional model and the initial three-dimensional model comprises:

[0023] A drawer mounting rail is generated on the drawer frame of the initial three-dimensional model, and the drawer three-dimensional model is assembled on the corresponding drawer mounting rail.

[0024] Optionally, the hand-drawn front view image is recognized by using a first recognition model, and the training step of the first recognition model includes:

[0025] A first training set is acquired, and the initial model is trained using the first training set to obtain an intermediate model; wherein the first training set includes hand-drawn training front views of cabinets of different users and labels corresponding to various parts in the hand-drawn training front views of cabinets;

[0026] Inputting a first test set into the intermediate model to obtain an output result of the intermediate model; comparing the output result with a label corresponding to the first test set;

[0027] In response to the recognition accuracy of the intermediate model reaching a threshold accuracy, the first recognition model is obtained.

[0028] Optionally, in step S2, the method further includes

[0029] The position of the molding front view is adjusted according to the positions of the hollow frame and the solid frame; wherein the hollow frame and the solid frame at different double ends are the top and the base line respectively.

[0030] Optionally, in step S4, generating a corresponding cabinet door three-dimensional model according to the cabinet door type label in the frame line and the position of the cabinet door type label includes:

[0031] According to the cabinet door type label, obtaining the type of the cabinet door;

[0032] Obtaining the size of the cabinet door according to the position and spanning area of ​​the cabinet door type label;

[0033] A corresponding three-dimensional model of the cabinet door is generated according to the type of the cabinet door and the size of the cabinet door.

[0034] Beneficial effects of the present invention: 1. By adding type labels to hand-drawn pictures, the present invention makes it easier for image recognition to extract features corresponding to different frames, and then it is easier to identify the types of frames, so that the corresponding three-dimensional models can be accurately constructed. The type label setting of the present invention can improve the accuracy of image recognition, avoid misidentification of frames with similar shapes, and avoid recognition deviations caused by the user's hand-drawn pictures being too sloppy. 2. The present invention can directly convert the front view into a three-dimensional model without the need for three views. Compared with the prior art, it greatly reduces the user's drawing time and improves the user experience. 3. The present invention can identify cabinet doors and drawers that need additional drawing and three-dimensional model generation through type labels, so that the two are generated together in the overall three-dimensional model generation, which can effectively reduce the modeling steps and increase the modeling efficiency. 4. The present invention uses a first recognition model for image recognition, and the first recognition model is trained using a training set that integrates various user data, which greatly improves the recognition accuracy. 5. The present invention can generate corresponding production and processing drawings based on the three-dimensional model to facilitate subsequent production and manufacturing.

[0035] In summary, the present invention can improve the accuracy of image recognition and build a three-dimensional model that meets the requirements. At the same time, the present invention can quickly convert the cabinet design sketch into a three-dimensional model. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a flowchart of a method for generating a three-dimensional model of a cabinet by hand-drawing based on AI image recognition provided by a specific embodiment of the present invention;

[0037] Figure 2 is a schematic diagram illustrating a hand-drawn picture and its corresponding markings provided in the first specific embodiment of the present invention;

[0038] Figure 3 It is a schematic diagram for explaining a hand-drawn picture and its corresponding marks provided in the second specific embodiment of the present invention. DETAILED DESCRIPTION

[0039] The present invention discloses a method for generating a three-dimensional model of a cabinet by hand-drawing based on AI image recognition. Those skilled in the art can refer to the content of this article and appropriately improve the technical details. It should be particularly noted that all similar replacements and modifications are obvious to those skilled in the art, and they are all considered to be included in the present invention. The method and application of the present invention have been described through preferred embodiments, and relevant personnel can obviously modify or appropriately change and combine the methods and applications described herein without departing from the content, spirit and scope of the present invention to implement and apply the technology of the present invention.

[0040] The applicant has discovered through research that: However, the current technology of constructing a cabinet three-dimensional model by intelligently recognizing hand-drawn sketches has errors in recognizing hand-drawn sketches, which makes the constructed three-dimensional model inaccurate. For example, it can identify the grid frame of the cabinet, but cannot accurately distinguish whether it is a storage compartment or a drawer compartment.

[0041] Therefore, the embodiment of the present invention provides a method for generating a three-dimensional model of a cabinet by hand-drawing based on AI image recognition, such as Figure 1 As shown, the method includes:

[0042] Step S1, obtaining a hand-drawn front view image corresponding to a first cabinet, performing image recognition on frame lines in the hand-drawn front view image; generating an initial front view of the first cabinet according to the recognized frame lines and the thickness of a plate material used to manufacture the first cabinet.

[0043] The initial front view includes multiple frames.

[0044] It should be noted that the rectangle formed by lines is referred to as a frame in the embodiment of the present invention. The hand-drawn front view image is drawn by a user through drawing software or hand-drawn scanned (photographed) and uploaded.

[0045] In this specific embodiment, the hand-drawn front view image is recognized by using a first recognition model, and the training steps of the first recognition model include:

[0046] A first training set is acquired, and the initial model is trained using the first training set to obtain an intermediate model; wherein the first training set includes hand-drawn training front views of cabinets of different users and labels corresponding to various parts in the hand-drawn training front views of cabinets;

[0047] Input the first test set into the intermediate model to obtain the output result of the intermediate model; compare the output result with the label corresponding to the first test set;

[0048] In response to the recognition accuracy of the intermediate model reaching a threshold accuracy, a first recognition model is obtained.

[0049] It should be noted that, by using a large number of hand-drawn pictures of different users for training, the first recognition model enhances the feature extraction capability and improves the recognition accuracy.

[0050] Step S2, image recognition of the type labels in the frame lines to obtain the types of various frames in the initial front view; according to the types of the frames, a frame of corresponding structure is generated in the initial front view to obtain a formed front view of the first cabinet.

[0051] Among them, the frame includes a fully enclosed frame, a semi-enclosed frame and a hollow frame. The semi-enclosed frame is divided into a storage cabinet frame and a drawer frame. The fully enclosed frame is divided into a hollow frame and a solid frame. The solid frame includes the baseboard of the first cabinet.

[0052] It should be noted that the frame in the actual cabinet may be the top blank, baseboards, drawers, storage compartments, etc., but these are difficult to express through hand-drawn front views, and thus difficult to identify, resulting in modeling deviations.

[0053] In a first specific embodiment, the hand-drawn elevation view can be Figure 2 As shown, in the frame, "○" represents a drawer frame, "+" represents a storage frame, "X" represents a hollow frame, oblique cross filling represents a hollow frame, and horizontal and vertical cross filling represents a solid frame. Figure 2 The embodiment uses a sliding door type cabinet door, so it is represented by left and right double-way arrows. The frame spanned by the two left and right double-way arrows is the horizontal coverage area of ​​the cabinet door, and the frame spanned by the up and down double-way arrows between the two left and right double-way arrows is the vertical coverage area of ​​the cabinet door. Figure 2 In the example, the four frames below are the cabinet door covering area. The oblique cross filling and the horizontal and vertical cross filling only need to be used for identification, and do not need to be very dense. The identification of the horizontal and vertical cross filling and the frame lines can depend on the density. This embodiment is for illustration only.

[0054] In a second specific embodiment, the hand-drawn front view can be as follows Figure 3 As shown, in the frame, "○" represents a drawer frame, "+" represents a storage frame, "X" represents a hollow frame, oblique cross filling represents a hollow frame, and horizontal and vertical cross filling represents a solid frame. Figure 2 In the embodiment, a swing-door type cabinet door is used, and thus “[” and “]” are used to represent it, and the frame between “[” and “]” is the area covered by the cabinet door. Figure 3 In the figure, the four frames below are the door covering area.

[0055] In this specific embodiment, in step S2, the method further includes

[0056] According to the positions of the hollow frame and the solid frame, the position of the molding front view is adjusted; wherein the hollow frame and the solid frame with different double ends are the top and the base line respectively.

[0057] It should be noted that the baseboard, generally called the baseboard or skirting in the industry, is generally 60-80mm high. The baseboard is solid to increase its bearing capacity, and because the thickness of the cabinet plate is generally specified, if the baseboard is to be solid, it is generally composed of several bars stacked together, so its thickness is an integer multiple of the plate thickness. This parameter can be used to adjust the baseboard thickness parameter input so that the thickness meets this requirement.

[0058] Step S3: perform three-dimensional stretching on the formed front view to construct an initial three-dimensional model of the first cabinet.

[0059] It should be noted that 3D stretching is generally performed according to a default ratio, which can be modified by the user later. Alternatively, the user can mark in a hand-drawn drawing and then identify the stretching ratio.

[0060] Step S4: Generate a corresponding three-dimensional model of the drawer according to the type label corresponding to the drawer frame; generate a corresponding three-dimensional model of the door according to the door type label within the frame line and the position of the door type label.

[0061] Among them, the types of cabinet doors include sliding door type and swing door type.

[0062] In this specific embodiment, in step S4, generating a corresponding cabinet door three-dimensional model according to the cabinet door type label in the frame line and the position of the cabinet door type label includes:

[0063] According to the cabinet door type label, obtain the cabinet door type;

[0064] According to the location and spanning area of ​​the door type label, the size of the door is obtained;

[0065] Generate a corresponding three-dimensional model of the cabinet door according to the type and size of the cabinet door.

[0066] It should be noted that the cabinet door type label of the present invention not only marks the type, but also standardizes the position and size, so that the function of the type label is more extensive, and the label is simplified while making its meaning clear.

[0067] Step S5: combine the initial three-dimensional model, the drawer three-dimensional model and the cabinet door three-dimensional model, and obtain a final three-dimensional model corresponding to the first cabinet according to the size parameters of the first cabinet.

[0068] In this specific embodiment, after step S5, the method further includes:

[0069] According to the structure of the final 3D model, corresponding installation ports and connectors are generated at various parts of the final 3D model; the installation ports and connectors are used to connect the plates at various parts during the actual manufacturing process of the first cabinet to form various frames of the first cabinet. The connectors include hinges, guide rails, etc.

[0070] It should be noted that, in the actual manufacturing process, the various parts need to be assembled and connected to form the first cabinet, so it is necessary to install ports and connectors to connect the various parts in an ideal state.

[0071] In this specific embodiment, before step S5, the method further includes:

[0072] Image recognition identifies the dimension markings of each frame line in the hand-drawn front view image to obtain the dimension parameters corresponding to the first cabinet.

[0073] In another specific embodiment, step S5 includes:

[0074] The initial three-dimensional model, the drawer three-dimensional model and the cabinet door three-dimensional model are combined to generate a default three-dimensional model; the size parameters of the default three-dimensional model are input to adjust the default size, and the final three-dimensional model corresponding to the first cabinet is generated according to the size parameters; wherein the default three-dimensional model has a default size. It should be noted that the size parameters can be identified by marking in the hand-drawn drawing in the early stage, or they can be input and adjusted after the rough model is generated. Both of these can be realized, which increases the application flexibility of the embodiments of the present invention.

[0075] In this specific embodiment, after step S5, the method further includes:

[0076] According to the final three-dimensional model, processing drawings of various panels for manufacturing the first cabinet corresponding to the final three-dimensional model are obtained.

[0077] It should be noted that from modeling to actual production, processing drawings are required. The embodiment of the present invention can generate drawings based on the model, which greatly facilitates the processing flow.

[0078] In this specific embodiment, in step S5, the initial three-dimensional model and the cabinet door three-dimensional model are combined, including:

[0079] According to the type of the cabinet door three-dimensional model, a first cabinet door installation component between the cabinet door and the initial three-dimensional model is obtained;

[0080] According to the size of the cabinet door, the number and distribution positions of the first cabinet door installation components are obtained;

[0081] The initial three-dimensional model and the cabinet door three-dimensional model are combined according to the first cabinet door installation components, the number of the first cabinet door installation components and the distribution positions.

[0082] In this specific embodiment, in step S5, combining the drawer three-dimensional model and the initial three-dimensional model includes:

[0083] A drawer mounting slide rail is generated on the drawer frame of the initial three-dimensional model, and the drawer three-dimensional model is assembled on the corresponding drawer mounting slide rail.

[0084] It should be noted that no matter the cabinet door installation components or the drawer installation rails, these components are extracted and generated. When the hand-drawn image is recognized and it is found that these components are needed, the components and component installation positions can be generated accordingly. At the same time, the number of components and component installation positions can also be determined according to the specifications of the cabinet door or drawer.

[0085] The embodiment of the present invention adds type labels to the hand-drawn picture, so that image recognition can more easily extract the features corresponding to different frames, and then it can more easily identify the type of frame, so that the corresponding three-dimensional model can be accurately constructed. The type label setting of the embodiment of the present invention can improve the accuracy of image recognition, avoid misidentification of frames with similar shapes, and avoid recognition deviation due to the user's hand-drawn picture being too sloppy.

[0086] The embodiment of the present invention can directly convert the front view into a three-dimensional model without the need for three views. Compared with the prior art, it greatly reduces the user's drawing time and improves the user experience.

[0087] The embodiment of the present invention can identify cabinet doors and drawers that require additional drawing and three-dimensional model generation through type tags, so that the two can be generated together in the overall three-dimensional model generation, which can effectively reduce modeling steps and increase modeling efficiency.

[0088] The embodiment of the present invention adopts a first recognition model to perform image recognition. The first recognition model is trained using a training set that integrates various user data, which greatly improves the recognition accuracy.

[0089] The embodiment of the present invention can generate corresponding production and processing drawings according to the three-dimensional model to facilitate subsequent production and manufacturing.

[0090] In summary, the embodiment of the present invention can improve the accuracy of image recognition and build a three-dimensional model that better meets the requirements. At the same time, the embodiment of the present invention can quickly convert the cabinet design sketch into a three-dimensional model.

[0091] It should be noted that, in this article, relational terms such as first and second, etc. are only used 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. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0092] Each embodiment in this specification is described in a related manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0093] The above are only preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention are included in the protection scope of the present invention.

Claims

1. A method for generating a three-dimensional model of a cabinet based on hand-drawn drawings based on AI image recognition, characterized in that: The method comprises: Step S1, obtaining a hand-drawn front view image corresponding to a first cabinet, and performing image recognition on frame lines in the hand-drawn front view image; generating an initial front view of the first cabinet according to the recognized frame lines and the thickness of a plate material used to manufacture the first cabinet; wherein the initial front view includes a plurality of frames; Step S2: image recognition of the type labels in the frame lines to obtain the types of the frames in the initial front view; according to the types of the frames, a frame of corresponding structure is generated in the initial front view to obtain a formed front view of the first cabinet; wherein the frames include fully enclosed frames, semi-enclosed frames and hollow frames, the semi-enclosed frames are divided into locker frames and drawer frames, the fully enclosed frames are divided into hollow frames and solid frames, and the solid frames include the baseboards of the first cabinet; Step S3, performing three-dimensional stretching on the formed front view to construct an initial three-dimensional model of the first cabinet; Step S4, generating a corresponding three-dimensional model of the drawer according to the type label corresponding to the drawer frame; generating a corresponding three-dimensional model of the cabinet door according to the cabinet door type label in the frame line and the position of the cabinet door type label; wherein the cabinet door types include sliding door type and swing door type; Step S5: combining the initial three-dimensional model, the drawer three-dimensional model and the cabinet door three-dimensional model, and obtaining a final three-dimensional model corresponding to the first cabinet according to the size parameters of the first cabinet.

2. The method for generating a three-dimensional cabinet model based on hand-drawn drawing based on AI image recognition according to claim 1 is characterized in that: After step S5, the method further includes: According to the structure of the final three-dimensional model, corresponding installation ports and connectors are generated at various parts of the final three-dimensional model; the installation ports and connectors are used to connect the panels at various parts during the actual manufacturing process of the first cabinet to form various frames of the first cabinet.

3. The method for generating a three-dimensional cabinet model based on hand-drawn drawing based on AI image recognition according to claim 1 is characterized in that: Before step S5, the method further includes: Image recognition is used to identify the dimension markings of each frame line in the hand-drawn front view image, so as to obtain the dimension parameters corresponding to the first cabinet.

4. The method for generating a three-dimensional cabinet model based on hand-drawn drawing based on AI image recognition according to claim 1 is characterized in that: The step S5 comprises: The initial three-dimensional model, the drawer three-dimensional model and the cabinet door three-dimensional model are combined to generate a default three-dimensional model; size parameters are input into the default three-dimensional model to adjust the default size, and a final three-dimensional model corresponding to the first cabinet is generated according to the size parameters; wherein the default three-dimensional model has a default size.

5. The method for generating a three-dimensional cabinet model based on hand-drawn drawing based on AI image recognition according to claim 1 is characterized in that: After step S5, the method further includes: According to the final three-dimensional model, processing drawings of various panels for manufacturing the first cabinet corresponding to the final three-dimensional model are obtained.

6. The method for generating a three-dimensional cabinet model based on hand-drawn drawing based on AI image recognition according to claim 1 is characterized in that: In the step S5, the initial three-dimensional model and the cabinet door three-dimensional model are combined, including: According to the type of the cabinet door three-dimensional model, obtaining a first cabinet door installation component between the cabinet door and the initial three-dimensional model; According to the size of the cabinet door, obtaining the number and distribution positions of the first cabinet door installation components; The initial three-dimensional model and the cabinet door three-dimensional model are combined according to the first cabinet door installation component, the number of the first cabinet door installation components and the distribution position.

7. The method for generating a three-dimensional cabinet model based on hand-drawn drawings based on AI image recognition according to claim 1 is characterized in that: In the step S5, combining the drawer three-dimensional model and the initial three-dimensional model comprises: A drawer mounting rail is generated on the drawer frame of the initial three-dimensional model, and the drawer three-dimensional model is assembled on the corresponding drawer mounting rail.

8. The method for generating a three-dimensional cabinet model based on hand-drawn drawing based on AI image recognition according to claim 1 is characterized in that: The hand-drawn frontal image is recognized by using a first recognition model, and the training steps of the first recognition model include: A first training set is acquired, and the initial model is trained using the first training set to obtain an intermediate model; wherein the first training set includes hand-drawn training front views of cabinets of different users and labels corresponding to various parts in the hand-drawn training front views of cabinets; Inputting a first test set into the intermediate model to obtain an output result of the intermediate model; comparing the output result with a label corresponding to the first test set; In response to the recognition accuracy of the intermediate model reaching a threshold accuracy, the first recognition model is obtained.

9. The method for generating a three-dimensional cabinet model based on hand-drawn drawings based on AI image recognition according to claim 1 is characterized in that: In step S2, the method further includes The position of the molding front view is adjusted according to the positions of the hollow frame and the solid frame; wherein the hollow frame and the solid frame at different double ends are the top and the base line respectively.

10. The method for generating a three-dimensional cabinet model based on hand-drawn drawings based on AI image recognition according to claim 1 is characterized in that: In step S4, generating a corresponding cabinet door three-dimensional model according to the cabinet door type label in the frame line and the position of the cabinet door type label includes: According to the cabinet door type label, obtaining the type of the cabinet door; Obtaining the size of the cabinet door according to the position and spanning area of ​​the cabinet door type label; A corresponding three-dimensional model of the cabinet door is generated according to the type of the cabinet door and the size of the cabinet door.

Images