Ceramic three-dimensional reconstruction method and system based on combination of laser radar and digital image

Abstract

The invention discloses a ceramic three-dimensional reconstruction method and system based on combination of a laser radar and a digital image, and relates to the technical field of three-dimensional reconstruction, and the ceramic three-dimensional reconstruction method comprises the steps: obtaining laser radar data and visual image data; based on the laser radar data, generating a three-dimensional grid model and a grid point normal vector of the target object; according to the normal vector of the grid point and a set texture data evaluation function, judging whether texture data corresponding to the grid point in the three-dimensional grid model is effective or not, and obtaining effective texture data which is not influenced by reflection; mapping the effective texture data to the surface of the three-dimensional grid model to generate a three-dimensional reconstruction result of the target object; the ceramic three-dimensional reconstruction system comprises a camera module, a laser radar module, an arc track, a translation rotating platform, a light source module and a three-dimensional reconstruction device. The three-dimensional reconstruction device executes the ceramic three-dimensional reconstruction method. According to the method and the system, the technical problems of incomplete three-dimensional model reconstruction and high repair difficulty caused by ceramic surface reflection are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional reconstruction, and in particular to a ceramic three-dimensional reconstruction method and system based on the combination of laser radar and digital images. Background Art

[0002] With the rapid development of image technology, the technology of 3D reconstruction of objects through single or multiple camera modules has gradually matured. Currently, the commonly used methods include multi-view based on feature point technology. Figure 3 Methods such as Structure from Motion (SFM), Neural Radiance Fields (Nerf), and Gaussian3D typically perform pose estimation on multi-view images after matching feature points to achieve 3D reconstruction of objects. However, these existing methods are less suitable for objects with smooth and reflective surfaces. For 3D reconstruction of ceramic objects, due to their smooth and reflective surfaces and lack of texture information, using these 3D reconstruction methods can result in a large number of damaged surfaces in the model, increasing the difficulty of later model repair and even leading to direct 3D reconstruction failure, making it impossible to truly restore the 3D shape and texture pattern of the ceramic object. Summary of the Invention

[0003] The main purpose of the present invention is to provide a ceramic three-dimensional reconstruction method and system based on the combination of lidar and digital images, aiming to solve the technical problems of incomplete three-dimensional model reconstruction and difficult repair caused by ceramic surface reflection in existing three-dimensional reconstruction methods and systems.

[0004] To achieve the above objectives, the present application provides a ceramic three-dimensional reconstruction method based on the combination of laser radar and digital images, the method comprising: Acquire lidar data and visual image data from multiple angles; Generate a 3D mesh model and mesh point normal vectors of the target object based on lidar data from multiple angles; Based on the grid point normal vector and a set texture data evaluation function, the validity of the texture data corresponding to the grid point in the three-dimensional grid model is determined. After removing the invalid texture data, valid texture data unaffected by reflections is obtained. The texture data is obtained based on visual image data at multiple angles. A grid point in the three-dimensional grid model corresponds to texture data at multiple different angles. The texture data evaluation function is used to describe the geometric distance relationship between the reflected light vector corresponding to the grid point in the three-dimensional grid model and the camera module position. The reflected light vector is calculated based on the grid point normal vector and a pre-known incident light vector. The effective texture data is mapped to the surface of the 3D mesh model to generate the 3D reconstruction result of the target object.

[0005] Optionally, generating a three-dimensional grid model and grid point normal vectors of the target object based on the lidar data at multiple angles includes: Perform coordinate transformation on the laser radar data at multiple angles to obtain the laser point cloud data of the target object; Based on the implicit surface reconstruction method and laser point cloud data, a laser 3D model of the target object is constructed; Based on the laser 3D model, a 3D grid model and grid point normal vectors of the target object are generated.

[0006] Optionally, performing coordinate transformation on the laser radar data at multiple angles to obtain laser point cloud data of the target object includes: Based on the set radar coordinate geometric relationship, the position data of the lidar module and the beam emission angle, the radar position coordinate system and the laser light line equation of the lidar module in the three-dimensional space coordinate system are obtained; the position data and beam emission angle of the lidar module are obtained based on the lidar data; According to the set rotation coordinate geometric relationship and the rotation angle of the target object, the rotation position coordinate system of the target object after rotation in the three-dimensional space coordinate system is obtained; the rotation angle of the target object is obtained based on the lidar data; According to the set translation coordinate geometric relationship and the translation distance of the target object, the translation position coordinate system of the target object after translation in the three-dimensional space coordinate system is obtained; the translation distance of the target object is obtained based on the lidar data; Establish the three-dimensional space coordinate system of the target object according to the radar position coordinate system, the laser light straight line equation, the rotation position coordinate system and the translation position coordinate system; Align the lidar data at multiple angles to the three-dimensional spatial coordinate system of the target object to obtain the laser point cloud data of the target object.

[0007] Optionally, the method for constructing the set texture data evaluation function includes: Obtaining the position data of the camera module and the preset light source position data corresponding to the visual image data at multiple angles; Align the camera module's position data and the preset light source position data to the target object's three-dimensional space coordinate system to obtain the camera position coordinates and the preset light source position coordinates corresponding to each visual image data; Obtaining an incident light vector corresponding to a grid point in the three-dimensional grid model according to the preset light source position coordinates and the grid point position coordinates corresponding to the grid point in the three-dimensional grid model; Based on the grid point normal vector corresponding to the grid point in the three-dimensional grid model and the incident light vector, a reflected light normal vector corresponding to the grid point in the three-dimensional grid model is obtained; Based on the incident light vector, the reflected light normal vector and the camera position coordinates, the relative distance value between the reflected light corresponding to the grid point in the three-dimensional grid model and the camera position coordinates is obtained; The set threshold parameter is compared with the relative distance value to obtain the texture data evaluation function corresponding to the visual image data at multiple angles.

[0008] Optionally, the determining whether texture data corresponding to a grid point in a three-dimensional grid model is valid based on the grid point normal vector and a set texture data evaluation function, and obtaining valid texture data not affected by reflection after removing invalid texture data, comprises: Based on the grid parameters of the three-dimensional grid model, the visual image data at multiple angles is divided into multiple texture data; According to the set texture data evaluation function, the geometric distance between the reflected light and the camera position of each texture data corresponding grid point is calculated to obtain the quality evaluation result; The texture data with a quality evaluation result less than zero is determined to be invalid texture data, otherwise it is determined to be valid texture data; Invalid texture data is removed to obtain valid texture data that is not affected by reflection.

[0009] Optionally, before acquiring the lidar data and visual image data at multiple angles, the method includes calibrating the lidar module and the camera module using a reference calibration plate.

[0010] In addition, to achieve the above-mentioned purpose, the present application also proposes a ceramic three-dimensional reconstruction system based on the combination of laser radar and digital images. The system includes a camera module, a laser radar module, an arc track, a translation and rotation platform, a light source module and a three-dimensional reconstruction device. The light source module is used to illuminate the target object. The translation and rotation platform is arranged in the circular arc track to drive the target object to perform rotational motion and / or translational motion. The camera module and the laser radar module are respectively arranged on the circular arc track and move along the circular arc track to adjust the shooting angle. The camera module and the laser radar module are respectively connected to the three-dimensional reconstruction device to collect visual image data and laser radar data at multiple angles. The three-dimensional reconstruction device is used to execute any of the ceramic three-dimensional reconstruction methods described above.

[0011] Furthermore, the laser radar module includes a laser and a receiving target surface, the receiving target surface is set as a local spherical surface, the laser emits a laser beam toward the center of the arc track based on a preset swing angle range to perform an up and down line scan of the target object, and the three-dimensional reconstruction device establishes a three-dimensional space coordinate system with the center of the arc track as the origin.

[0012] Furthermore, the translation and rotation platform includes a rotating platform and a translation platform. The translation platform is used to drive the target object to translate horizontally. The rotating platform is arranged inside the translation platform to drive the target object to rotate 360°. The three-dimensional reconstruction device establishes a three-dimensional space coordinate system with the rotation axis of the rotating platform as the Z direction and the translation direction of the translation platform as the X direction.

[0013] Furthermore, there are two groups of camera modules, which are respectively arranged on both sides of the laser radar module.

[0014] The present invention adopts the above technical solution to achieve the following effects: The ceramic 3D reconstruction method of the present invention combines laser radar technology with digital imaging technology, constructs a high-precision ceramic 3D model through laser radar technology, and then uses texture mapping technology to map effective texture data to the surface of the ceramic 3D model, thereby solving the problems of incomplete 3D model reconstruction and difficulty in repair caused by reflection on the ceramic surface, and effectively and realistically restores the 3D shape and texture pattern of the ceramic object; the ceramic 3D reconstruction system of the present invention includes the above-mentioned ceramic 3D reconstruction method, combines laser radar technology with digital imaging technology, and effectively and realistically restores the 3D shape and texture pattern of the ceramic object. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A schematic structural diagram of a ceramic 3D reconstruction system according to an embodiment of the present invention is shown; Figure 2 A schematic diagram showing a flow chart of a ceramic 3D reconstruction method according to an embodiment of the present invention; Figure 3 A schematic structural diagram of a three-dimensional grid model according to an embodiment of the present invention is shown; Figure 4 A schematic diagram showing the positions of a camera module, a light source module, and a three-dimensional grid model according to an embodiment of the present invention; Figure 5 A schematic diagram showing the conversion relationship of a radar position coordinate system in a ceramic 3D reconstruction method according to an embodiment of the present invention; Figure 6 A schematic diagram showing the conversion relationship of a rotation position coordinate system in a ceramic 3D reconstruction method according to an embodiment of the present invention; Figure 7 A schematic diagram showing the positional transformation relationship of the three-dimensional space coordinate system of a target object in a ceramic three-dimensional reconstruction method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0016] Various exemplary embodiments, features, and aspects of the present invention will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise indicated.

[0017] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0018] The term "and / or" herein simply describes an association relationship between associated objects, indicating that three relationships can exist. For example, "A and / or B" can represent the existence of three situations: A alone, A and B simultaneously, and B alone. Furthermore, the term "at least one" herein refers to any combination of at least two of any one or more of a plurality of items. For example, "at least one of A, B, and C" can represent any one or more elements selected from the set consisting of A, B, and C.

[0019] It should be understood that the terms "first," "second," and "third," etc. in the claims, description, and drawings of the present invention are used to distinguish different objects, rather than to describe a specific order. The terms "comprise" and "comprising" as used in the description and claims of the present invention indicate the presence of the described features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or combinations thereof.

[0020] In addition, numerous specific details are provided in the following detailed description to better illustrate the present invention. Those skilled in the art will appreciate that the present invention may be practiced without certain specific details. In some instances, methods, means, components, and circuits well known to those skilled in the art have not been described in detail in order to highlight the main points of the present invention.

[0021] For example, Figure 1As shown, the ceramic three-dimensional reconstruction system provided by the embodiment of the present invention includes an arc track 1, a translation and rotation platform, a laser radar module, camera modules 4T and 4B, a base 6, a light source module, a three-dimensional reconstruction device 7 and a target object 8. The laser radar module includes a laser 4L and a receiving target surface 5, and the translation and rotation platform includes a rotating platform 3 and a translation platform 2. The arc track 1 is fixedly mounted on the base 6. The laser 4L, camera modules 4T and 4B are fixedly connected to the receiving target surface 5 and are mounted on the arc track 1. They can move along the arc track 1 on the arc track 1. The translation platform 2 is mounted on the base 6 and realizes horizontal movement in a single direction through a slide rail. The rotating platform 3 is mounted on the translation platform 2 to realize single-direction rotation. The target object 8 (i.e., ceramic) is placed on the rotating platform 3. The light source module can be set below the camera module (such as Figure 4 Position H as shown is used to illuminate the target object. The receiving target surface 5 is a partially spherical surface with multiple arrayed fiber optic bundles distributed across it. Signals are transmitted via the fiber optic bundles to the receiver and then to the 3D reconstruction device. Laser 4L scans up and down within the plane of the circular track 1. Simultaneously, laser 4L and camera modules 4T and 4B move along the circular track 1, adjusting the angle between the laser beam from laser 4L and the target object 8, allowing more reflection points to fall on the partially spherical receiving target surface 5. Combined with the translation platform 2 and the rotation platform 3, the entire target object can be scanned. The coordinated operation of multiple movements ensures that the laser light is captured by the receiver after reflection from the target object. It is understood that when the target object 8 is placed on the rotation platform 3, the rotation platform 3 and the translation platform 2 drive the target object 8 to rotate or translate to scan the target object according to the set scanning route. After a single scan, the position of the lidar module and camera module on the circular track can be adjusted to change the shooting angle and perform multiple scans, thereby obtaining lidar data and visual image data from multiple angles.

[0022] After the laser radar module and the camera module collect the laser radar data and visual image data at multiple angles, they are transmitted to the 3D reconstruction device 7. The 3D reconstruction device 7 performs data preprocessing and 3D reconstruction processing on the laser radar data and visual image data at multiple angles, and executes the 3D reconstruction method steps as follows: Figure 2 As shown, the method includes the following steps S100~S400.

[0023] Step S100: Acquire lidar data and visual image data at multiple angles.

[0024] It should be noted that before step S100, the laser radar module and the camera module can be calibrated using a reference calibration plate. The reference calibration plate is used to accurately calculate the spatial relationship between the laser radar module and the camera module, which is beneficial to improving the accuracy of reconstruction. The laser radar data of the embodiment of the present invention includes the position data of the laser radar module, the swing angle data of the laser beam, the rotation angle data of the rotating platform, and the translation angle data of the translation platform, etc., which are fused with the original signal data in the subsequent data processing to obtain laser point cloud data, and then construct a three-dimensional model. The position data of the laser radar module can be obtained based on the position of the laser radar module on the circular track 1, which can include the position information of the laser 4 and the receiving target surface 5.

[0025] Step S200: Generate a three-dimensional grid model and grid point normal vectors of the target object based on the lidar data at multiple angles.

[0026] In this embodiment of the present invention, after performing data preprocessing, registration, and fusion on the LiDAR data from multiple angles, a 3D model of the target object is constructed. The 3D model of the target object is then meshed to obtain the 3D mesh model and mesh point normals of the target object. For example, step S200 includes the following steps S210 through S230.

[0027] Step S210 , performing coordinate transformation on the laser radar data at multiple angles to obtain laser point cloud data of the target object.

[0028] That is, the laser radar module has its own coordinate system when scanning at different positions. Multi-view data fusion requires all data to be aligned and optimized in the same coordinate system. Through coordinate transformation, data from different perspectives can be aligned to the three-dimensional space coordinate system of the same target object to obtain the aligned laser point cloud data.

[0029] Illustratively, the method for establishing a three-dimensional space coordinate system of a target object according to an embodiment of the present invention may include the following steps S211 to S214 .

[0030] Step S211, according to the set radar coordinate geometric relationship, the position data of the laser radar module and the beam emission angle, obtain the radar position coordinate system and the laser light line equation of the laser radar module in the three-dimensional space coordinate system; the position data and beam emission angle of the laser radar module are obtained based on the laser radar data.

[0031] For example, see Figure 5, the laser 4 emits a laser beam toward the center of the circular arc track based on a preset swing angle range to perform an up and down line scan of the target object, and the three-dimensional reconstruction device establishes a three-dimensional space coordinate system with the center of the circular arc track as the origin. With the center of the circular arc track 1 as the center, the direction perpendicular to the circular surface of the circular arc track 1 is the x-axis, and the horizontal direction of the circular surface is the y-axis to establish a rectangular coordinate system. After installation, the laser 4 in the laser radar module emits a laser beam AQ toward the center of the circle. The laser radar module moves along the track in the circular arc track 1, and its movement range is the upper half of the circular arc of the circular arc track 1. .

[0032] The radar coordinate geometric relationship is set as Figure 5 As shown, assuming that the laser 4 is located at point A, the radius of the arc track 1 is , the angle between the connecting line AO ​​and the z-axis is , then the coordinates of the laser 4 in the rectangular coordinate system xyz are , then the radar position coordinate system is expressed as: ; ; ; The laser beam AQ can be swept up and down in the yoz plane with the connecting line AO ​​as the center line, and its swing angle is , specifies the upward swing angle of the laser beam AQ is positive, otherwise it is negative. Then the straight line equation of the laser beam AQ in the yoz plane is: ; in, .

[0033] Step S212, obtaining the rotation position coordinate system of the target object after rotation in the three-dimensional space coordinate system according to the set rotation coordinate geometric relationship and the rotation angle of the target object; the rotation angle of the target object is obtained based on the laser radar data.

[0034] For example, the rotation coordinate geometry can be found in Figure 6 , the target object 8 is placed on the rotating platform 3, and the rotating platform 3 realizes the target object 8 to rotate 360° around the z axis. The coordinates of any point P on the target object 8 in the rectangular coordinate system xyz are , the coordinates of point P Written in polar coordinate form, ,angle When rotating around the z axis When the angle is (positive according to the right-hand rule), the coordinates after rotation are , then the rotation position coordinate system is expressed as: ; ; ; Step S213, obtaining the translation position coordinate system of the target object after translation in the three-dimensional space coordinate system according to the set translation coordinate geometric relationship and the translation distance of the target object; the translation distance of the target object is obtained based on the laser radar data.

[0035] For example, the translation coordinate geometry relationship can be found in Figure 6 , the translation platform 2 can realize the movement in the x direction, and the translation platform 2 moves After (i.e. translation distance , it is stipulated that along the positive x direction is positive, vice versa is negative), translation The coordinates after , then the translation position coordinate system is expressed as: ; ; ; Step S214 , establishing a three-dimensional space coordinate system of the target object according to the radar position coordinate system, the laser light line equation, the rotation position coordinate system and the translation position coordinate system.

[0036] For example, the radar position coordinate system, the laser line equation, the rotation position coordinate system and the translation position coordinate system obtained in the above steps S211 to S213 are integrated to obtain the three-dimensional space coordinate system of the target object. Figure 7 , assuming that the laser beam AP is reflected by the surface of the target object 8 and then received by the receiving target surface 5, the laser reaches the position of point B on the receiving target surface, and its coordinates are , according to the recorded laser beam APB time , we can know that: ; Where, is the distance to the AP, is the distance in PB, It's the speed of light.

[0037] From the above, the coordinates of point P are calculated based on the geometric relationship Expressed as: ; ; ; The normal vector of point P is: ; Therefore, the coordinates of point P in the polar coordinate system can be calculated as: ; ; Then, it is converted into the coordinates of the XYZ coordinate system (that is, the three-dimensional space coordinate system of the target object) as follows: ; ; ; Step S215 , aligning the laser radar data at multiple angles to the three-dimensional space coordinate system of the target object to obtain laser point cloud data of the target object.

[0038] It can be understood that the three-dimensional reconstruction device establishes a three-dimensional space coordinate system with the center of the arc track as the origin, the rotation axis of the rotating platform as the Z direction, and the translation direction of the translation platform as the X direction. The three-dimensional space coordinate system of the target object is established based on the three-dimensional space coordinate system, and the lidar data at multiple angles are aligned to the three-dimensional space coordinate system of the target object to ensure the fusion of multi-angle data and generate an accurate and consistent three-dimensional model.

[0039] Step S220 : constructing a laser three-dimensional model of the target object based on the implicit surface reconstruction method and the laser point cloud data.

[0040] For example, after obtaining the laser point cloud data and normal vector of the target object, an implicit surface reconstruction method based on the Poisson algorithm can be used for three-dimensional reconstruction. The vertex results of the output mesh are smoother, and due to the global solution, the watertightness of the mesh can be guaranteed.

[0041] Step S230 : generating a three-dimensional grid model and grid point normal vectors of the target object based on the laser three-dimensional model.

[0042] Exemplarily, the laser three-dimensional model obtained in step 220 is solved by solving the Poisson equation to construct a smooth implicit function, and then the grid is extracted to generate the final three-dimensional grid model. The grid points can be interpolation points and / or grid vertices within the grid, and the grid normal vector can be calculated according to the above normal vector expression. It can be understood that the three-dimensional grid model in the embodiment of the present invention is a three-dimensional grid model with a normal vector (such as Figure 3 ), which helps the subsequent texture mapping results to be smoother.

[0043] Step S300, based on the grid point normal vector and the set texture data evaluation function, determines whether the texture data corresponding to the grid point in the three-dimensional grid model is valid, and after removing the invalid texture data, obtains the valid texture data that is not affected by the reflection; wherein: the texture data is obtained based on the visual image data at multiple angles; a grid point in the three-dimensional grid model corresponds to texture data at multiple different angles; the texture data evaluation function is used to describe the geometric distance relationship between the reflected light vector corresponding to the grid point in the three-dimensional grid model and the camera module position, and the reflected light vector is calculated based on the grid point normal vector and the pre-known incident light vector.

[0044] It should be noted that, because the images of ceramic objects under illumination will reflect light, the reflective areas of the images taken at different angles are different. Through the visual image data taken at multiple angles, a large amount of image data of the surface of ceramic objects has been obtained. In other words, each grid point in the three-dimensional grid model corresponds to texture data at multiple different angles, and the texture data at some angles has reflective conditions. The validity of the texture data is determined according to the set texture data evaluation function. The corresponding reflected light vector can be calculated based on the known grid point normal vector and the incident light vector. Based on the geometric distance relationship between the reflected light vector and the camera position, it can be determined whether the reflected light directly enters the camera lens. In other words, it can be determined whether the captured image data has reflective conditions at the current grid point. The reflective texture data of the grid point, i.e., invalid texture data, is removed to obtain valid texture data that is not affected by the reflection.

[0045] An exemplary method for constructing a texture data evaluation function includes: Step S310: Obtaining the camera module position data and preset light source position data corresponding to the visual image data at multiple angles. It is understood that each recorded visual image data may be tagged with the camera module position data and the preset light source position data, and the preset light source position data may also be pre-determined based on the installation position of the light source module.

[0046] In step S320 , the position data of the camera module and the preset light source position data are aligned to the three-dimensional space coordinate system of the target object to obtain the camera position coordinates and the preset light source position coordinates corresponding to each visual image data.

[0047] Step S330 , obtaining an incident light vector corresponding to a grid point in the three-dimensional grid model according to the preset light source position coordinates and the grid point position coordinates corresponding to the grid point in the three-dimensional grid model.

[0048] Step S340 , obtaining a reflected light normal vector corresponding to the grid point in the three-dimensional grid model based on the grid point normal vector corresponding to the grid point in the three-dimensional grid model and the incident light vector.

[0049] Step S350 , based on the incident light vector, the reflected light normal vector and the camera position coordinates, obtain the relative distance value between the reflected light corresponding to the grid point in the three-dimensional grid model and the camera position coordinates.

[0050] Step S360 : Compare the set threshold parameter with the relative distance value to obtain texture data evaluation functions corresponding to the visual image data at multiple angles.

[0051] by Figure 4 Let’s take an example. Assuming that the camera position and light source position have been aligned to the three-dimensional space coordinate system, the coordinates of the current target object grid point P in the coordinate system are , the normal vector is , the light source position H is , the camera position A is , then the incident light vector is The calculated normal vector of the reflected light is: ; The distance between the camera position and the reflected light is expressed as: ; because , which includes the camera position and the position of the reflected light If we expand the equation and write the formula, it will be very large. ; ; ; Therefore, the above expression can be used to determine the geometric distance relationship between the reflected light vector corresponding to the grid point in the three-dimensional grid model and the camera module position. In other words, based on the geometric distance relationship between the reflected light vector and the camera position, it can be determined whether the reflected light directly enters the camera lens. The texture data evaluation function is set to be expressed as: ,in, It is the set threshold parameter used to determine whether the reflected light data is valid, that is, whether the reflected light directly enters the camera. , it means that the texture data is invalid and the texture data may have reflections. Otherwise, it means that the texture data is valid. The valid texture data is retained and the texture data affected by reflections is removed.

[0052] S400: Mapping the effective texture data to the surface of the three-dimensional mesh model to generate a three-dimensional reconstruction result of the target object.

[0053] For example, after constructing a three-dimensional network model using lidar technology, an embodiment of the present invention establishes a mapping relationship between the three-dimensional network model and the texture map. The three-dimensional mesh model can use UV mapping to align the surface of the three-dimensional model with the two-dimensional texture image, and map the texture data to the surface of the three-dimensional mesh model to generate a three-dimensional reconstruction result of the target object.

[0054] In summary, the ceramic three-dimensional reconstruction method of the present invention combines lidar technology with digital imaging technology, constructs a high-precision ceramic three-dimensional model through lidar technology, and then uses texture mapping technology to map effective texture data to the surface of the ceramic three-dimensional model, thereby solving the problems of incomplete three-dimensional model reconstruction and difficulty in repair caused by reflection on the ceramic surface, and effectively and realistically restoring the three-dimensional shape and texture pattern of ceramic objects. First, a three-dimensional mesh model and mesh point normal vectors are generated using lidar data from multiple angles. Since the images of ceramic objects obtained under light will have reflections, the reflection areas of the images obtained at different angles are different. A large amount of image data of the surface of ceramic objects has been obtained through visual image data taken at multiple angles, that is, each grid point in the three-dimensional mesh model corresponds to texture data at multiple angles, and the texture data at some angles has reflections. Then, the validity of the texture data is judged according to the set texture data evaluation function. The corresponding reflected light vector can be calculated based on the known grid point normal vector and the incident light vector. According to the geometric distance relationship between the reflected light vector and the camera position, it can be judged whether the reflected light directly enters the camera lens, that is, whether the captured image data has reflections at the current grid point position. The reflective texture data of the grid point, that is, invalid texture data, is removed to obtain valid texture data that is not affected by reflections. Finally, the valid texture data is mapped to the surface of the three-dimensional mesh model to generate the three-dimensional reconstruction result of the target object.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. A ceramic 3D reconstruction method based on the combination of laser radar and digital images, characterized in that: The method comprises: Acquire lidar data and visual image data from multiple angles; Generate a 3D mesh model and mesh point normal vectors of the target object based on lidar data from multiple angles; Based on the grid point normal vector and a set texture data evaluation function, the validity of the texture data corresponding to the grid point in the three-dimensional grid model is determined. After removing the invalid texture data, valid texture data unaffected by reflections is obtained. The texture data is obtained based on visual image data at multiple angles. A grid point in the three-dimensional grid model corresponds to texture data at multiple different angles. The texture data evaluation function is used to describe the geometric distance relationship between the reflected light vector corresponding to the grid point in the three-dimensional grid model and the camera module position. The reflected light vector is calculated based on the grid point normal vector and a pre-known incident light vector. The effective texture data is mapped to the surface of the 3D mesh model to generate the 3D reconstruction result of the target object.

2. The ceramic three-dimensional reconstruction method according to claim 1, characterized in that: The method of generating a three-dimensional grid model and grid point normal vectors of a target object based on the laser radar data at multiple angles includes: Perform coordinate transformation on the laser radar data at multiple angles to obtain the laser point cloud data of the target object; Based on the implicit surface reconstruction method and laser point cloud data, a laser 3D model of the target object is constructed; Based on the laser 3D model, a 3D grid model and grid point normal vectors of the target object are generated.

3. The ceramic three-dimensional reconstruction method according to claim 1, characterized in that: The coordinate transformation of the laser radar data at multiple angles to obtain laser point cloud data of the target object includes: Based on the set radar coordinate geometric relationship, the position data of the lidar module and the beam emission angle, the radar position coordinate system and the laser light line equation of the lidar module in the three-dimensional space coordinate system are obtained; the position data and beam emission angle of the lidar module are obtained based on the lidar data; According to the set rotation coordinate geometric relationship and the rotation angle of the target object, the rotation position coordinate system of the target object after rotation in the three-dimensional space coordinate system is obtained; the rotation angle of the target object is obtained based on the lidar data; According to the set translation coordinate geometric relationship and the translation distance of the target object, the translation position coordinate system of the target object after translation in the three-dimensional space coordinate system is obtained; the translation distance of the target object is obtained based on the lidar data; Establish the three-dimensional space coordinate system of the target object according to the radar position coordinate system, the laser light straight line equation, the rotation position coordinate system and the translation position coordinate system; Align the lidar data at multiple angles to the three-dimensional spatial coordinate system of the target object to obtain the laser point cloud data of the target object.

4. The ceramic three-dimensional reconstruction method according to claim 1, characterized in that: The construction method of the set texture data evaluation function includes: Obtaining the position data of the camera module and the preset light source position data corresponding to the visual image data at multiple angles; Align the camera module's position data and the preset light source position data to the target object's three-dimensional space coordinate system to obtain the camera position coordinates and the preset light source position coordinates corresponding to each visual image data; Obtaining an incident light vector corresponding to a grid point in the three-dimensional grid model according to the preset light source position coordinates and the grid point position coordinates corresponding to the grid point in the three-dimensional grid model; Based on the grid point normal vector corresponding to the grid point in the three-dimensional grid model and the incident light vector, a reflected light normal vector corresponding to the grid point in the three-dimensional grid model is obtained; Based on the incident light vector, the reflected light normal vector and the camera position coordinates, the relative distance value between the reflected light corresponding to the grid point in the three-dimensional grid model and the camera position coordinates is obtained; The set threshold parameter is compared with the relative distance value to obtain the texture data evaluation function corresponding to the visual image data at multiple angles.

5. The ceramic three-dimensional reconstruction method according to claim 1, characterized in that: The method comprises: determining whether the texture data corresponding to the grid point in the three-dimensional grid model is valid based on the grid point normal vector and the set texture data evaluation function, and obtaining the valid texture data not affected by the reflection after removing the invalid texture data; Based on the grid parameters of the three-dimensional grid model, the visual image data at multiple angles is divided into multiple texture data; According to the set texture data evaluation function, the geometric distance between the reflected light and the camera position of each texture data corresponding grid point is calculated to obtain the quality evaluation result; The texture data with a quality evaluation result less than zero is determined as invalid texture data, otherwise it is determined as valid texture data; Invalid texture data is removed to obtain valid texture data that is not affected by reflection.

6. The ceramic three-dimensional reconstruction method according to claim 1, characterized in that: Before acquiring the laser radar data and visual image data at multiple angles, the method includes: Use the reference calibration plate to calibrate the lidar module and camera module.

7. A ceramic 3D reconstruction system based on the combination of laser radar and digital images, characterized in that: The system includes a camera module, a laser radar module, an arc track, a translation and rotation platform, a light source module and a three-dimensional reconstruction device. The light source module is used to illuminate the target object. The translation and rotation platform is arranged in the arc track to drive the target object to perform rotational motion and / or translational motion. The camera module and the laser radar module are correspondingly arranged on the arc track and move along the arc track to adjust the shooting angle. The camera module and the laser radar module are respectively connected to the three-dimensional reconstruction device to collect visual image data and laser radar data at multiple angles. The three-dimensional reconstruction device is used to execute the ceramic three-dimensional reconstruction method as described in any one of claims 1 to 6.

8. The system according to claim 7, characterized in that The laser radar module includes a laser and a receiving target surface. The receiving target surface is set as a local spherical surface. The laser emits a laser beam toward the center of the arc track based on a preset swing angle range to scan the target object up and down. The three-dimensional reconstruction device establishes a three-dimensional space coordinate system with the center of the arc track as the origin.

9. The system according to claim 7, wherein: The translation and rotation platform includes a rotating platform and a translation platform. The translation platform is used to drive the target object to translate horizontally. The rotating platform is arranged inside the translation platform to drive the target object to rotate 360°. The 3D reconstruction device establishes a 3D space coordinate system with the rotation axis of the rotating platform as the Z direction and the translation direction of the translation platform as the X direction.

10. The system according to claim 7, wherein: There are two groups of camera modules, which are respectively located on both sides of the lidar module.

Images