Lightweight modeling method, system and equipment of mesh model and medium

Abstract

The invention discloses a lightweight modeling method, system and device for a mesh model and a medium, and the method comprises the steps: carrying out the oblique photography of a target object, and obtaining three-dimensional image data; identifying the three-dimensional image data to obtain block attributes of the target object; constructing a corresponding material model based on the block attributes; constructing a mesh model contour of the target object based on the three-dimensional image data; and filling the material model into the mesh model contour to obtain a mesh live-action model of the target object. The method solves the problem that an existing mesh model is only composed of geometric surfaces, attribute information of a real object cannot be known based on the mesh model, and the authenticity of the model is low.

Description

Technical Field

[0001] The present invention relates to the technical field of mesh model modeling, and in particular to a lightweight modeling method, system, equipment and medium for a mesh model. Background Art

[0002] The mesh model is a grid structure composed of many connected points and lines, which can accurately present the three-dimensional shape and surface details of the object.

[0003] A mesh model is a 3D model made of grids that can create various shapes and structures by connecting many small polygons together. Each polygon is called a face, and the faces are connected by shared vertices to form a mesh. This model has high flexibility and fine detail expression capabilities, and is often used in 3D design and simulation in various fields.

[0004] However, the existing mesh model is only composed of geometric surfaces, and it is impossible to understand the attribute information of the real object based on the mesh model, resulting in low model authenticity.

[0005] Therefore, how to solve the problem that the existing mesh model cannot understand the attribute information of the real object based on the mesh model, resulting in low model authenticity has become a technical problem that technical personnel in this field urgently need to solve. Summary of the invention

[0006] In view of the above-mentioned defects of the prior art, the present invention provides a lightweight modeling method, system, equipment and medium for a mesh model, the purpose of which is to overcome the problem that the existing mesh model is composed only of geometric surfaces and the attribute information of the real object cannot be understood based on the mesh model, resulting in low model authenticity.

[0007] To achieve the above object, the present invention discloses a lightweight modeling method of a mesh model, comprising the following steps:

[0008] Step 1: Obliquely photograph the target object to obtain three-dimensional image data of the target object;

[0009] Step 2: Identify the three-dimensional image data to obtain the block attributes of the target object;

[0010] Step 3: construct a corresponding material model based on the block attributes;

[0011] Step 4: constructing a mesh model contour of the target object based on the three-dimensional image data;

[0012] Step 5: Fill the material model into the mesh model outline to obtain a mesh real-scene model of the target object.

[0013] Preferably, step 1 is as follows:

[0014] Step 1.1, performing oblique photography on the target object to obtain initial image data;

[0015] Step 1.2, preprocessing the initial image data, removing objects other than the target object from the initial image data, and obtaining the three-dimensional image data of the target object;

[0016] The preprocessing includes noise reduction processing and de-additive processing, through which noise information and irrelevant information in the initial image data are eliminated to restore useful real information.

[0017] Preferably, step 2 is as follows:

[0018] Step 2.1, dividing the types of image elements belonging to the target object in the three-dimensional image data to obtain a plurality of element blocks;

[0019] Step 2.2: using a preset attribute recognition model to identify all the element blocks, and obtaining the attributes of all the element blocks, which are the block attributes of the target object;

[0020] In step 2.1, the image element types include point elements, line elements, surface elements, text elements, line elements, dot elements, cell elements and / or geometric figure elements;

[0021] Step 2.2 is as follows:

[0022] Step 2.2.1, using the preset attribute recognition model to extract features of element blocks, to obtain feature information of all the element blocks;

[0023] Step 2.2.2: Analyze all the feature information to obtain the attributes of all the element blocks.

[0024] Preferably, the feature information includes shape feature, color feature, surface texture feature, physical state feature, and transparency feature, and the block attributes include shape, color, surface texture, physical state, and transparency;

[0025] The physical state characteristics include solid state characteristics, liquid state characteristics and gas state characteristics;

[0026] The physical state includes solid, liquid and gas;

[0027] The block attributes include shape, color, surface texture, physical state, and transparency.

[0028] Preferably, step 3 is as follows:

[0029] Step 3.1, based on the color, surface texture, physical state and transparency of the block attributes, find the corresponding basic material model in the basic model library;

[0030] Step 3.2: Cut the basic material model based on the shape of the block attribute to obtain a material model corresponding to the block attribute.

[0031] Preferably, step 4 is as follows:

[0032] Step 4.1, determining the contour boundary position of the target object based on the top view, left view, right view and rear view of the target object in the three-dimensional image data;

[0033] Step 4.2, using the elevation normal of the target object in the three-dimensional image data as the main elevation direction, contour drawing is performed on the contour boundary position to obtain the mesh model contour of the target object;

[0034] Step 4.2 is as follows:

[0035] Step 4.2.1, identifying the three-dimensional image data, and determining the real space position corresponding to the material model in the target object;

[0036] Step 4.2.2, obtaining a position correspondence between the target object and the mesh model contour, and based on the position correspondence, searching for a target position corresponding to the real space position in the mesh model contour;

[0037] Step 4.2.3: Fill the material model into the corresponding target position in the mesh model outline to obtain the mesh real-scene model of the target object.

[0038] The present invention also provides a lightweight modeling system of a mesh model, which is used to run the lightweight modeling method of the mesh model, specifically comprising:

[0039] The photography module is used to perform oblique photography of the target object to obtain three-dimensional image data, specifically:

[0040] Performing oblique photography on the target object to obtain the initial image data;

[0041] Preprocessing the initial image data to remove objects other than the target object in the initial image data to obtain the three-dimensional image data;

[0042] The recognition module is used to recognize the three-dimensional image data and obtain the block attributes of the target object, specifically:

[0043] Dividing the types of image elements of the target object in the three-dimensional image data to obtain a plurality of element blocks;

[0044] Identify the element block using a preset attribute recognition model to obtain a block attribute of the element block;

[0045] Alternatively, a preset attribute recognition model is used to extract features of the element block to obtain feature information of the element block;

[0046] Analyzing the characteristic information to obtain the block attribute of the element block;

[0047] The first construction module is used to construct a corresponding material model based on the block attributes, specifically:

[0048] Based on the color, surface texture, physical state, and transparency of the block attributes, a corresponding basic material model is searched in a basic model library;

[0049] Cutting the basic material model based on the shape of the block attribute to obtain a material model corresponding to the block attribute;

[0050] The second construction module is used to construct a mesh model contour of the target object based on the three-dimensional image data, specifically:

[0051] Determine the contour boundary position of the target object based on the top view, left view, right view and rear view of the target object in the three-dimensional image data;

[0052] Using the elevation normal of the target object in the three-dimensional image data as the main elevation direction, contour drawing is performed on the contour boundary position to obtain the mesh model contour of the target object;

[0053] The obtaining module is used to fill the material model into the mesh model outline to obtain the mesh real scene model of the target object, specifically:

[0054] Recognize the three-dimensional image data and determine the real space position corresponding to the material model in the target object;

[0055] Obtain the position correspondence between the target object and the mesh model contour, and based on the position correspondence, find the target position corresponding to the real space position in the mesh model contour;

[0056] Fill the material model to the target position in the model outline to obtain a mesh real-scene model of the target object.

[0057] The present invention also provides a lightweight modeling device for a mesh model, comprising a memory, a processor, and a program stored in the memory and running on the processor;

[0058] When the processor executes the program, the steps of the lightweight modeling method of the above-mentioned mesh model are implemented.

[0059] The present invention also provides a lightweight modeling medium of a mesh model, which is a computer-readable storage medium;

[0060] The computer-readable storage medium stores instructions, and when the instructions are executed on a terminal device, the terminal device executes the steps of the lightweight modeling method of the mesh model.

[0061] Beneficial effects of the present invention:

[0062] The present invention obtains three-dimensional image data by performing oblique photography on the target object, recognizes the three-dimensional image data, obtains the block attributes of the target object, and constructs a corresponding material model based on the block attributes. The material model has the attribute information of the corresponding block attributes, so that the authenticity of the material model is improved.

[0063] The present invention constructs a mesh model contour of a target object based on three-dimensional image data, and fills a material model into the mesh model contour to obtain a mesh real-scene model of the target object.

[0064] The present invention does not need to perform targeted drawing of real-scene details on the mesh model. By directly filling the material model with block attributes into the mesh model outline of the target object, the attributes of the target object can be introduced into the mesh real-scene model, and the mesh real-scene model with attribute information can be directly obtained. This not only reduces the drawing operations of the real-scene model and realizes lightweight modeling, but also directly understands the attribute information of the target object through the mesh real-scene model, thereby improving the authenticity of the mesh real-scene model.

[0065] The concept, specific structure and technical effects of the present invention will be further described below in conjunction with the accompanying drawings to fully understand the purpose, characteristics and effects of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 A schematic flow chart of a lightweight modeling method for a mesh model in one embodiment of the present invention is shown.

[0067] Figure 2 A schematic diagram showing a flow chart of a lightweight modeling method of a mesh model in another embodiment of the present invention

[0068] Figure 3A schematic structural diagram of a lightweight modeling system of a mesh model in one embodiment of the present invention is shown. DETAILED DESCRIPTION

[0069] Example 1

[0070] like Figure 1 As shown in the figure, the lightweight modeling method of the mesh model includes the following steps:

[0071] Step 1: Obliquely photograph the target object to obtain three-dimensional image data of the target object;

[0072] Step 2: Identify the three-dimensional image data to obtain the block attributes of the target object;

[0073] Step 3: Build the corresponding material model based on the block attributes;

[0074] Step 4: construct a mesh model outline of the target object based on the three-dimensional image data;

[0075] Step 5: Fill the material model into the mesh model outline to obtain the mesh real-scene model of the target object.

[0076] The present invention obtains three-dimensional image data by performing oblique photography on the target object, recognizes the three-dimensional image data, obtains the block attributes of the target object, and constructs a corresponding material model based on the block attributes. The material model has the attribute information of the corresponding block attributes, so that the authenticity of the material model is improved.

[0077] The present invention constructs a mesh model contour of a target object based on three-dimensional image data, and fills a material model into the mesh model contour to obtain a mesh real-scene model of the target object.

[0078] Through the above-mentioned technical means, the present invention does not need to carry out targeted drawing of real-scene details of the mesh model. The material model with block attributes is directly filled into the mesh model outline of the target object, so that the attributes of the target object can be introduced into the mesh real-scene model, and the mesh real-scene model with attribute information can be directly obtained. This not only reduces the drawing operation of the real-scene model and realizes lightweight modeling, but also can directly understand the attribute information of the target object through the mesh real-scene model, thereby improving the authenticity of the mesh real-scene model.

[0079] In some embodiments, step 1 is as follows:

[0080] Step 1.1, take oblique photography of the target object to obtain initial image data;

[0081] Step 1.2, preprocessing the initial image data, removing objects other than the target object from the initial image data, and obtaining three-dimensional image data of the target object;

[0082] The present invention utilizes the oblique photography technology to perform oblique photography on the target object.

[0083] Oblique photography technology is a high-tech technology developed in the international field of photogrammetry in the past decade. This technology obtains rich high-resolution textures of building tops and side views by synchronously collecting images from one vertical, four oblique, and five different perspectives.

[0084] Oblique photogrammetry technology comprehensively perceives complex scenes in a large-scale, high-precision, and high-definition manner. The data results generated by efficient data acquisition equipment and professional data processing processes intuitively reflect the appearance, position, height and other attributes of the ground objects, providing guarantees for realistic effects and surveying and mapping-level accuracy.

[0085] It can also effectively improve the production efficiency of the model. The modeling of a small or medium-sized city that would take one or two years to complete using manual modeling can now be completed in three to five months using oblique photography modeling, greatly reducing the economic and time costs of three-dimensional model data collection.

[0086] Preprocessing includes noise reduction and de-additive processing, which can eliminate noise information and irrelevant information in the initial image data and restore useful real information.

[0087] In practical applications, preprocessing includes noise reduction and de-impurity processing.

[0088] The initial image data obtained by oblique photography of the target object can truly reflect the ground conditions and surface texture information of the target object's environment. Therefore, the initial image data is preprocessed to eliminate noise information and irrelevant information in the initial image data, restore useful real information, enhance the detectability of relevant information of the target object, and simplify the data to the maximum extent, so as to obtain three-dimensional image data that can truly reflect the target object's conditions and surface texture information, so as to facilitate the subsequent more accurate construction of a material model that matches the target object, thereby improving the accuracy of the obtained mesh real-scene model.

[0089] In some embodiments, step 2 is as follows:

[0090] Step 2.1, dividing the types of image elements belonging to the target object in the three-dimensional image data to obtain a plurality of element blocks;

[0091] Step 2.2: Use the preset attribute recognition model to identify all element blocks, and obtain the attributes of all element blocks, which are the block attributes of the target object;

[0092] In practical applications, by dividing the types of image elements of the target object in the three-dimensional image data, multiple element blocks that can constitute the target object are obtained, and the element blocks are identified using a preset attribute recognition model to obtain the block attributes of the element blocks, which facilitates the subsequent construction of a material model that can constitute the target object based on the block attributes, thereby improving the accuracy of the obtained mesh real-scene model.

[0093] In step 2.1, the types of image elements include point elements, line elements, surface elements, text elements, line elements, dot elements, cell elements and / or geometric figure elements.

[0094] Step 2.2 is as follows:

[0095] Step 2.2.1, extracting features from element blocks using a preset attribute recognition model to obtain feature information of all element blocks;

[0096] Step 2.2.2: Analyze all feature information to obtain the attributes of all element blocks.

[0097] In practical applications, a preset attribute recognition model is used to extract features of element blocks to obtain feature information that can reflect the element features of the element blocks.

[0098] The characteristic information is then analyzed to obtain block attributes that can reflect the element characteristics of the element block, so as to facilitate the subsequent construction of a material model with attribute information of the corresponding block attributes based on the block attributes, thereby improving the authenticity of the material model and further improving the accuracy of the obtained mesh real-scene model.

[0099] In some embodiments, the feature information includes shape features, color features, surface texture features, physical state features, and transparency features, and the block attributes include shape, color, surface texture, physical state, and transparency;

[0100] Physical state characteristics include solid state characteristics, liquid state characteristics and gas state characteristics;

[0101] Physical states include solid, liquid, and gas;

[0102] Block attributes include shape, color, surface texture, physical state, and transparency.

[0103] In some embodiments, step 3 is as follows:

[0104] Step 3.1, based on the color, surface texture, physical state and transparency of the block attributes, find the corresponding basic material model in the basic model library;

[0105] Step 3.2: Cut the basic material model based on the shape of the block attributes to obtain the material model corresponding to the block attributes.

[0106] In practical applications, basic material models of different colors, different surface textures, different physical states, and different transparencies are established in advance and stored in the basic model library, so that when it is necessary to build a corresponding material model based on block attributes, the corresponding basic material model can be directly found in the basic model library based on the color, surface texture, physical state, and transparency of the block attributes;

[0107] Then, the basic material model is cut based on the shape of the block attributes to obtain the material model corresponding to the block attributes, without drawing and constructing, thus reducing the drawing operation process of the material model, thereby realizing the lightweight model construction of the material model.

[0108] In some embodiments, step 4 is as follows:

[0109] Step 4.1, determining the contour boundary position of the target object based on the top view, left view, right view and rear view of the target object in the three-dimensional image data;

[0110] Step 4.2, using the elevation normal of the target object in the three-dimensional image data as the main elevation direction, contour drawing is performed on the contour boundary position to obtain the mesh model contour of the target object;

[0111] In practical applications, the top view, left view, right view and rear view of the target object contained in the three-dimensional image data can clearly understand the contour boundary of the target object from various angles, thereby determining the contour boundary position of the target object, and then using the facade normal of the target object in the three-dimensional image data as the main facade direction, the contour boundary position is contoured to obtain a mesh model contour matching the target object, which is convenient for subsequently filling the lightweight material model with block attributes into the mesh model contour of the target object, and directly obtaining a mesh real-scene model with attribute information. This not only realizes lightweight modeling of the mesh real-scene model, but also directly understands the attribute information of the target object through the mesh real-scene model, thereby improving the authenticity of the mesh real-scene model.

[0112] Step 4.2 is as follows:

[0113] Step 4.2.1, recognize the three-dimensional image data and determine the real space position corresponding to the material model in the target object;

[0114] Step 4.2.2, obtaining the position correspondence between the target object and the mesh model contour, and based on the position correspondence, searching for the target position corresponding to the real space position in the mesh model contour;

[0115] Step 4.2.3: Fill the material model into the corresponding target position in the mesh model contour to obtain the mesh real-scene model of the target object.

[0116] In practical applications, based on the positional correspondence between the target object and the mesh model contour, and the real-space position corresponding to the identified material model in the target object, the target position corresponding to the real-space position is searched in the mesh model contour, and the material model is filled into the target position in the model contour to obtain the mesh real-scene model of the target object.

[0117] Through the above-mentioned technical means, this embodiment can automatically fill the material model into the model outline by constructing a position correspondence between the target object and the mesh model outline. There is no need to manually identify the position where the material model needs to be filled in the mesh model outline, nor is there a need to manually fill the material model into the model outline. This can not only reduce labor costs, but also improve the modeling efficiency of the mesh real-scene model.

[0118] Example 2

[0119] Figure 2 As another exemplary embodiment, a flowchart of a lightweight modeling method using a provided mesh model is provided, wherein the specific steps include:

[0120] Step S201, taking oblique photography of the target building to obtain initial image data;

[0121] Step S202, pre-processing the initial image data to remove other buildings except the target building in the initial image data to obtain three-dimensional image data;

[0122] Step S203, dividing the image element types of the target building in the three-dimensional image data into multiple element blocks;

[0123] Step S204, extracting features of the element block using a preset attribute recognition model to obtain feature information of the element block;

[0124] Step S205, analyzing the characteristic information to obtain block attributes of the element block, the block attributes include shape, color, surface texture, physical state, and transparency;

[0125] Step S206, based on the color, surface texture, physical state, and transparency of the block attributes, find the corresponding basic material model in the basic model library;

[0126] Step S207, cutting the basic material model based on the shape of the block attribute to obtain a material model corresponding to the block attribute;

[0127] Step S208, determining the outline boundary position of the target building based on the top view angle, left view angle, right view angle and rear view angle of the target building in the three-dimensional image data;

[0128] Step S209, using the facade normal of the target building in the three-dimensional image data as the main facade direction, contour drawing is performed on the contour boundary position to obtain the mesh model contour of the target building;

[0129] Step S210, identifying the three-dimensional image data to determine the real space position corresponding to the material model in the target building;

[0130] Step S210, obtaining a position correspondence relationship between the target building and the mesh model outline, and based on the position correspondence relationship, searching for a target position corresponding to a real space position in the mesh model outline;

[0131] Step S212, filling the material model into the target position in the model outline to obtain a mesh real-scene model of the target building.

[0132] In this embodiment, the element blocks corresponding to the target building include window blocks, wall blocks, eaves blocks, stair blocks, roof truss blocks, beam blocks, slab blocks, column blocks, and the like.

[0133] First, by taking oblique photography of the target building and preprocessing the initial image data obtained by the oblique photography, the noise information and other irrelevant information in the image are eliminated, the useful real information is restored, the detectability of the relevant information of the target building is enhanced, and the data is simplified to the maximum extent, so as to obtain three-dimensional image data that can truly reflect the situation of the target building and the surface texture information of the object.

[0134] Secondly, the types of image elements of the target building in the three-dimensional image data are divided to obtain multiple element blocks that can constitute the target building, and the element blocks are identified using a preset attribute recognition model to obtain block attributes that can reflect the element characteristics of the element blocks.

[0135] Thirdly, based on the block attributes, the corresponding basic material model is found in the basic model library, and the basic material model is cut to obtain the material model corresponding to the block attributes. The material model has the attribute information of the corresponding block attributes, which improves the authenticity of the material model. There is no need to draw the material model corresponding to the block attributes, which reduces the drawing operation process of the material model, thereby enabling the lightweight model construction of the material model.

[0136] Based on the three-dimensional image data, the contour boundary of the target building can be clearly understood from all angles, so as to determine the contour boundary position of the target building, and use the facade normal of the target building in the three-dimensional image data as the main facade direction to draw the contour of the contour boundary position, and obtain the mesh model contour that matches the target building.

[0137] Finally, by constructing the position correspondence between the target building and the mesh model outline, the material model can be automatically filled into the mesh model outline to obtain the mesh real-scene model of the target building.

[0138] Through the above-mentioned technical means, there is no need to carry out targeted drawing of real-scene details of the mesh model. The attributes of the target building can be introduced into the mesh real-scene model by directly filling the material model with block attributes into the mesh model outline of the target building. The mesh real-scene model with attribute information can be directly obtained, which not only reduces the drawing operation of the real-scene model and realizes lightweight modeling, but also directly understands the attribute information of the target building through the mesh real-scene model, thereby improving the authenticity of the mesh real-scene model.

[0139] Moreover, through the method of this embodiment, there is no need to manually identify the position where the material model needs to be filled in the mesh model outline, nor is there any need to manually fill the material model into the model outline, which can not only reduce labor costs but also improve the modeling efficiency of the mesh real-scene model.

[0140] like Figure 3 As shown, the present invention also provides a lightweight modeling system of a mesh model, which is used to run the lightweight modeling method of the above mesh model, specifically including:

[0141] The photography module is used to perform oblique photography of the target object to obtain three-dimensional image data, specifically:

[0142] Perform oblique photography on the target object to obtain initial image data;

[0143] Preprocessing the initial image data to remove objects other than the target object in the initial image data to obtain three-dimensional image data;

[0144] The recognition module is used to recognize the three-dimensional image data and obtain the block attributes of the target object, specifically:

[0145] Dividing the types of image elements of the target object in the three-dimensional image data to obtain a plurality of element blocks;

[0146] Using a preset attribute recognition model to identify the element block, and obtaining the block attributes of the element block;

[0147] Alternatively, a preset attribute recognition model is used to extract features of the element block to obtain feature information of the element block;

[0148] Analyze the characteristic information to obtain the block attributes of the element block;

[0149] The first construction module is used to construct the corresponding material model based on the block attributes, specifically:

[0150] Based on the block attributes’ color, surface texture, physical state, and transparency, find the corresponding basic material model in the basic model library;

[0151] Cut the basic material model based on the shape of the block attributes to obtain the material model corresponding to the block attributes;

[0152] The second construction module is used to construct the mesh model contour of the target object based on the three-dimensional image data, specifically:

[0153] Determine the contour boundary position of the target object based on the top view, left view, right view and rear view of the target object in the three-dimensional image data;

[0154] The vertical normal of the target object in the three-dimensional image data is used as the main vertical direction, and the contour boundary position is contoured to obtain the mesh model contour of the target object;

[0155] The module is used to fill the material model into the mesh model outline to obtain the mesh real scene model of the target object, specifically:

[0156] Recognize the three-dimensional image data and determine the real space position corresponding to the material model in the target object;

[0157] Obtain the position correspondence between the target object and the mesh model contour, and based on the position correspondence, find the target position corresponding to the real space position in the mesh model contour;

[0158] Fill the material model to the target position in the model outline to obtain a mesh real-scene model of the target object.

[0159] In practical applications, the target object is photographed obliquely through the photography module to obtain three-dimensional image data.

[0160] The recognition module is used to recognize the three-dimensional image data to obtain the block attributes of the target object, and the first construction module is used to construct a corresponding material model based on the block attributes. The material model has attribute information of the corresponding block attributes, thereby improving the authenticity of the material model.

[0161] The mesh model contour of the target object is constructed based on the three-dimensional image data using the second construction module, and the material model is filled into the mesh model contour using the obtaining module to obtain the mesh real scene model of the target object.

[0162] Through the above technical means, there is no need to carry out targeted drawing of real-scene details of the mesh model. The attributes of the target object can be introduced into the mesh real-scene model by directly filling the material model with block attributes into the mesh model outline of the target object. The mesh real-scene model with attribute information can be directly obtained. This not only reduces the drawing operations of the real-scene model and realizes lightweight modeling, but also directly understands the attribute information of the target object through the mesh real-scene model, thereby improving the authenticity of the mesh real-scene model.

[0163] It should be noted that the lightweight modeling system of the mesh model provided in the above embodiment and the lightweight modeling method of the mesh model provided in the above embodiment belong to the same concept, wherein the specific manner in which each module and unit performs the operation has been described in detail in the method embodiment and will not be repeated here. In practical applications, the lightweight modeling system of the mesh model provided in the above embodiment can distribute the above functions to different functional modules as needed, that is, divide the internal structure of the system into different functional modules to complete all or part of the functions described above, and this is not limited here.

[0164] The present invention also provides a lightweight modeling device for a mesh model, comprising a memory, a processor, and a program stored in the memory and running on the processor;

[0165] The processor executes the program to implement the steps of the lightweight modeling method of the above mesh model.

[0166] In practical applications, the computing device can be a computer, and correspondingly, its program is computer software. The above-mentioned parameters and steps in a computing device of the present application can refer to the parameters and steps in the embodiment of the lightweight modeling method of the mesh model above, and will not be repeated here.

[0167] The present invention also provides a lightweight modeling medium of a mesh model, which is a computer-readable storage medium;

[0168] Instructions are stored in the computer-readable storage medium. When the instructions are executed on the terminal device, the terminal device executes the steps of the lightweight modeling method of the above-mentioned mesh model.

[0169] In practical applications, the computer-readable storage medium may be a transient computer-readable storage medium or a non-transitory computer-readable storage medium.

[0170] The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A lightweight modeling method for mesh models; characterized in that: The steps include: Step 1: Obliquely photograph the target object to obtain three-dimensional image data of the target object; Step 2: Identify the three-dimensional image data to obtain the block attributes of the target object; Step 3: construct a corresponding material model based on the block attributes; Step 4: constructing a mesh model contour of the target object based on the three-dimensional image data; Step 5: Fill the material model into the mesh model outline to obtain a mesh real-scene model of the target object.

2. The lightweight modeling method of the mesh model according to claim 1, characterized in that: Step 1 is as follows: Step 1.1, performing oblique photography on the target object to obtain initial image data; Step 1.2, preprocessing the initial image data, removing objects other than the target object from the initial image data, and obtaining the three-dimensional image data of the target object; The preprocessing includes noise reduction processing and de-additive processing, through which noise information and irrelevant information in the initial image data are eliminated to restore useful real information.

3. The lightweight modeling method of the mesh model according to claim 2, characterized in that: Step 2 is as follows: Step 2.1, dividing the types of image elements belonging to the target object in the three-dimensional image data to obtain a plurality of element blocks; Step 2.2: using a preset attribute recognition model to identify all the element blocks, and obtaining the attributes of all the element blocks, which are the block attributes of the target object; In step 2.1, the image element types include point elements, line elements, surface elements, text elements, line elements, dot elements, cell elements and / or geometric figure elements; Step 2.2 is as follows: Step 2.2.1, using the preset attribute recognition model to extract features of element blocks, to obtain feature information of all the element blocks; Step 2.2.2: Analyze all the feature information to obtain the attributes of all the element blocks.

4. The lightweight modeling method, system, device and medium of the mesh model according to claim 3, characterized in that: The feature information includes shape feature, color feature, surface texture feature, physical state feature, and transparency feature, and the block attributes include shape, color, surface texture, physical state, and transparency; The physical state characteristics include solid state characteristics, liquid state characteristics and gas state characteristics; The physical state includes solid, liquid and gas; The block attributes include shape, color, surface texture, physical state, and transparency.

5. The lightweight modeling method, system, device and medium of the mesh model according to claim 4, characterized in that: Step 3 is as follows: Step 3.1, based on the color, surface texture, physical state and transparency of the block attributes, find the corresponding basic material model in the basic model library; Step 3.2: Cut the basic material model based on the shape of the block attribute to obtain a material model corresponding to the block attribute.

6. The lightweight modeling method of the mesh model according to claim 5, characterized in that: Step 4 is as follows: Step 4.1, determining the contour boundary position of the target object based on the top view, left view, right view and rear view of the target object in the three-dimensional image data; Step 4.2, using the elevation normal of the target object in the three-dimensional image data as the main elevation direction, contour drawing is performed on the contour boundary position to obtain the mesh model contour of the target object; Step 4.2 is as follows: Step 4.2.1, identifying the three-dimensional image data, and determining the real space position corresponding to the material model in the target object; Step 4.2.2, obtaining a position correspondence between the target object and the mesh model contour, and based on the position correspondence, searching for a target position corresponding to the real space position in the mesh model contour; Step 4.2.3: Fill the material model into the corresponding target position in the mesh model outline to obtain the mesh real-scene model of the target object.

7. Lightweight modeling system of mesh model, characterized by: A lightweight modeling method for running a mesh model as claimed in any one of claims 1 to 6, specifically comprising: The photography module is used to perform oblique photography of the target object to obtain three-dimensional image data, specifically: Performing oblique photography on the target object to obtain the initial image data; Preprocessing the initial image data to remove objects other than the target object in the initial image data to obtain the three-dimensional image data; The recognition module is used to recognize the three-dimensional image data and obtain the block attributes of the target object, specifically: Dividing the types of image elements of the target object in the three-dimensional image data to obtain a plurality of element blocks; Identify the element block using a preset attribute recognition model to obtain a block attribute of the element block; Alternatively, a preset attribute recognition model is used to extract features of the element block to obtain feature information of the element block; Analyzing the characteristic information to obtain the block attribute of the element block; The first construction module is used to construct a corresponding material model based on the block attributes, specifically: Based on the color, surface texture, physical state, and transparency of the block attributes, a corresponding basic material model is searched in a basic model library; Cutting the basic material model based on the shape of the block attribute to obtain a material model corresponding to the block attribute; The second construction module is used to construct a mesh model contour of the target object based on the three-dimensional image data, specifically: Determine the contour boundary position of the target object based on the top view, left view, right view and rear view of the target object in the three-dimensional image data; Using the elevation normal of the target object in the three-dimensional image data as the main elevation direction, contour drawing is performed on the contour boundary position to obtain the mesh model contour of the target object; The obtaining module is used to fill the material model into the mesh model outline to obtain the mesh real scene model of the target object, specifically: Recognize the three-dimensional image data and determine the real space position corresponding to the material model in the target object; Obtain the position correspondence between the target object and the mesh model contour, and based on the position correspondence, find the target position corresponding to the real space position in the mesh model contour; Fill the material model to the target position in the model outline to obtain a mesh real-scene model of the target object.

8. Lightweight modeling equipment for mesh model, characterized in that: comprising a memory, a processor, and a program stored in the memory and running on the processor; When the processor executes the program, the steps of the lightweight modeling method of the mesh model as described in any one of claims 1 to 6 are implemented.

9. Lightweight modeling medium for mesh model, characterized by: A computer-readable storage medium; The computer-readable storage medium stores instructions, and when the instructions are executed on a terminal device, the terminal device executes the steps of the lightweight modeling method of a mesh model as described in any one of claims 1 to 6.

Images