A flexible 3D printing forming method and printing equipment based on 3D vision

Abstract

The present invention relates to the technical field of 3D printing, and in particular to a flexible 3D printing forming method and printing device based on 3D vision. The method comprises: obtaining a three-dimensional model file of a target object, calculating a voxel model and a point cloud model of the target object; obtaining a three-dimensional size ratio R of an axis-aligned bounding box of the voxel model, and generating a bottom-largest filling voxel unit a on a bottom surface of the voxel model according to the three-dimensional size ratio R; segmenting the voxel model according to the filling voxel unit a to generate a preset number of sub-voxel units; decomposing each sub-voxel unit until a sub-voxel with a bounding box size smaller than a preset spatial size exists in each decomposed sub-voxel, numbering all the sub-voxels to obtain a model decomposition sequence tree, the model decomposition sequence tree including a plurality of sub-voxel sequences; printing the target object according to the model decomposition sequence tree and the point cloud model, and improving the stability and reliability of the printing forming process.

Description

Technical Field

[0001] The present invention relates to the field of 3D printing technology, and in particular to a flexible 3D printing forming method and printing equipment based on 3D vision. Background Art

[0002] 3D printing technology, also known as "additive manufacturing technology", is an emerging manufacturing technology that uses digital models as the basis to stack materials layer by layer to create entities. It embodies the close integration of information network technology with advanced materials technology and digital manufacturing technology. Among them, the fused deposition modeling (FDM) process is widely used in the consumer field due to its simple structure and high cost performance. For process types such as wire arc additive manufacturing (WAAM) and laser near-net shaping (LENS), they are widely used in shipbuilding, aviation, aerospace, automobile, construction and other fields due to their advantages such as fast forming speed, large forming size, high material utilization rate and high strength of metal printed parts.

[0003] The 3D printing data processing of the above different process types can be summarized into two strictly related processing processes: (1) the digital model dimensionality reduction and discretization process; (2) the forming material accumulation manufacturing process. Among them, the dimensionality reduction and discretization process of the digital model is defined as follows: first, the 3D digital model (such as STL model, OBJ model, etc.) used to represent three-dimensional geometric information is input into the 3D printing system, and then the 3D digital model is sliced ​​layer by layer to obtain 2D contour data (SLC model, CLI model, etc.) used to represent two-dimensional geometric information. This link completes the dimensionality reduction and discretization process from the three-dimensional digital model to the two-dimensional digital model; finally, the 2D contour digital model is path planned to obtain 1D processing path data used to represent one-dimensional geometric information. This link completes the dimensionality reduction and discretization process from the two-dimensional digital model to the one-dimensional digital model; and the subsequent forming material accumulation manufacturing process can be regarded as a strict inverse of the previous process, and its process definition is as follows: the 3D printing system strictly follows the reverse process of "dimensionality reduction and discretization" based on the 1D processing path data obtained by dimensionality reduction and discretization, and controls the forming material to perform material accumulation manufacturing in a "reverse discretization" forming method from point to line, line to surface, and surface to body. Since the previous process is full of various strict and rigid digital model discretization algorithms, such as two-dimensional discretization slicing algorithms, one-dimensional discretization path planning algorithms, etc., and the subsequent process is strictly dependent on the previous discretization process, it ensures that 3D printing technology can realize the printing and manufacturing of any complex three-dimensional model into a three-dimensional printed entity.

[0004] However, in the above-mentioned printing method, on the one hand, the subsequent forming material accumulation process cannot be carried out strictly in accordance with the previous digital model dimensionality reduction and discretization process. For example, processing interference caused by mechanical structure factors of the equipment during the forming process may lead to printing failure; on the other hand, it is necessary to ensure that the functions of each equipment component of the entire 3D printing system are normal and in the best condition, so as to avoid forming defects caused by problems with equipment components; therefore, in the existing 3D printing forming method, the subsequent forming material accumulation manufacturing process needs to strictly and rigidly rely on the previous digital model dimensionality reduction and discretization process. Such a strongly dependent 3D printing forming method reduces the stability and reliability of the printing forming process. Summary of the Invention

[0005] Based on this, it is necessary to provide a flexible 3D printing forming method and printing equipment based on 3D vision to address the above technical problems, in order to solve the existing 3D printing forming method. The subsequent forming material accumulation manufacturing process needs to strictly and rigidly rely on the previous digital model dimensionality reduction discrete process, which seriously limits the forming capability and application scenarios of 3D printing.

[0006] In a first aspect, an embodiment of the present invention provides a flexible 3D printing method based on 3D vision, the method comprising:

[0007] Step S100: obtaining a three-dimensional model file of a target object, performing topological reconstruction on the three-dimensional model file to obtain a patch model of the target object; and calculating a voxel model and a point cloud model of the target object based on the patch model;

[0008] Step S200: obtaining a three-dimensional size ratio R of the axis-aligned bounding box of the voxel model, and generating a bottom-largest filling voxel unit a on the bottom surface of the voxel model according to the three-dimensional size ratio R;

[0009] Step S300: segmenting the voxel model according to the filled voxel unit a to generate a preset number of sub-voxel units;

[0010] Step S400: Decomposing each of the subvoxel units until a subvoxel with a bounding box size smaller than a preset spatial size exists among the decomposed subvoxels, numbering all the subvoxels to obtain a model decomposition sequence tree, wherein the model decomposition sequence tree includes a plurality of subvoxel sequences;

[0011] Step S500: Printing the target object according to the model decomposition sequence tree and the point cloud model to obtain a model of the target object.

[0012] Optionally, generating a bottom-largest filling voxel unit a on the bottom surface of the voxel model according to the three-dimensional size ratio R includes:

[0013] According to the voxel pattern shape of the bottom layer of the voxel model, obtaining a voxel point located at the center of the voxel pattern shape of the bottom layer of the voxel model as a seed voxel;

[0014] According to the three-dimensional size ratio R, the seed voxel is proportionally expanded until the expanded voxel contacts the outermost voxel of the voxel model;

[0015] The expanded voxel is moved one voxel point in any of the four directions of the horizontal plane. If the expanded voxel does not intersect with the outermost voxels of the voxel model, the expanded voxel is further expanded according to the three-dimensional size ratio R until the expanded voxel intersects with the outermost voxels of the voxel model by moving one voxel point in any of the four directions of the horizontal plane.

[0016] Optionally, segmenting the voxel model according to the filled voxel unit a to generate a preset number of sub-voxel units includes:

[0017] The plane where the remaining five surfaces of the filling voxel unit a excluding the bottom surface are located is used as a segmentation plane, and the voxel model is divided into eighteen sub-voxel units, and the filling voxel unit a is used as the final sub-voxel.

[0018] Optionally, decomposing each of the subvoxel units includes:

[0019] For any subvoxel unit m excluding the filling voxel unit a, obtaining a three-dimensional size ratio Q of the axis-aligned bounding box of the subvoxel unit m, generating a filling voxel unit b with the largest bottom on the bottom surface of the subvoxel unit m according to the three-dimensional size ratio Q, and using the filling voxel unit b as the final subvoxel;

[0020] The subvoxel unit m is recursively decomposed to generate several final subvoxels.

[0021] Optionally, all subvoxels are numbered to obtain a model decomposition sequence tree, including:

[0022] Obtain a first value of the top surface of each final sub-voxel on the Z axis and a second value of the bottom surface on the Z axis;

[0023] Taking each final sub-voxel with the smallest second value as the bottom sub-voxel, and numbering each bottom sub-voxel;

[0024] For any bottom-level subvoxel l, associate the bottom-level subvoxels whose second values ​​are the same as the first value of the bottom-level subvoxel l with the bottom-level subvoxel l among the unnumbered bottom-level subvoxels to obtain the subvoxel sequence corresponding to the bottom-level subvoxel l;

[0025] The model decomposition sequence tree is obtained according to the sub-voxel sequence corresponding to each underlying sub-voxel.

[0026] Optionally, printing the target object according to the model decomposition sequence tree and the point cloud model includes:

[0027] Decomposing the model sequence tree, obtaining a rough model of the target object, and printing the inner filling area of ​​the rough model layer by layer from bottom to top, setting the printing thickness of each layer to f;

[0028] Calculating the printing height h=i*f of the i-th layer of the rough model, obtaining a slice contour where the slice position of the i-th layer intersects with the rough model of the target object based on the printing height of the i-th layer, and performing fill printing on the i-th layer based on the slice contour;

[0029] Obtain a current point cloud model of the i-th layer of the rough model, compare the current point cloud model with the point cloud model of the target object to obtain a remaining area of ​​the i-th layer, and print the remaining area of ​​the i-th layer according to a preset printing thickness d, where d = f / k, and k is an integer greater than 1.

[0030] Optionally, printing the remaining area of ​​the i-th layer according to a preset printing thickness d includes:

[0031] The nozzle of the printer is positioned at the inner edge of the remaining area through the 3D vision module, and the remaining area of ​​the i-th layer is printed k times according to the flood filling algorithm until the thickness of the remaining area of ​​the i-th layer is equal to the preset printing thickness f.

[0032] Optionally, the method further includes:

[0033] Calculate the height h1=i of the rough model after printing max *f, based on the height h1 and the point cloud model of the target object, calculate the area to be printed above the height h1; i max is the maximum number of layers corresponding to the rough model when f is the printing thickness of each layer, i max *f≤p<(i max +1)*f, p is the height of the rough model;

[0034] The area to be printed is printed according to the preset printing thickness d to obtain a model of the target object.

[0035] Optionally, printing the to-be-printed area according to the preset printing thickness d includes:

[0036] Acquire a point cloud model of a model with a current printing height of h1, and compare the point cloud model of the model with the current printing height of h1 with the point cloud model of the target object to obtain a set U of the to-be-printed area;

[0037] Calculate the current printing height h2 = h1 + j*d, where j is the number of times the area to be printed is printed. Traverse the different area positions in the set U and use the flood filling algorithm to print each area with a height of h2 in the set U until the current printing height h2 reaches the maximum height of the point cloud model of the target object.

[0038] In a second aspect, an embodiment of the present invention provides a printing device, comprising a processor, a memory, and a printing device program stored in the memory and runnable on the processor, wherein when the processor executes the printing device program, the flexible 3D printing forming method based on 3D vision as described in the first aspect is implemented.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] The flexible 3D printing method based on 3D vision of the present invention weakens the strict and rigid dependency between the subsequent forming and manufacturing process and the previous model discrete process in the current 3D printing process during execution. During the 3D printing process, the point cloud model of the printed entity is collected by 3D vision and continuously compared and corrected with the original three-dimensional model, so that the three-dimensional entity in the actual printing can continuously and gradually approach the input three-dimensional model space geometric dimensions, thereby improving the stability and reliability of the printing process. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0042] Figure 1 1 is a flow chart of a flexible 3D printing method based on 3D vision provided in one embodiment of the present invention;

[0043] Figure 2 is a schematic diagram of voxel expansion provided in one embodiment of the present invention;

[0044] Figure 3 is a schematic diagram of a segmented subvoxel unit provided in one embodiment of the present invention;

[0045] Figure 42 is a schematic diagram of a voxel model decomposition provided in one embodiment of the present invention. DETAILED DESCRIPTION

[0046] In the following description, specific details such as particular system structures and techniques are provided for purposes of illustration and not limitation to facilitate a thorough understanding of the embodiments of the present invention. However, it will be apparent to those skilled in the art that the present invention may be practiced in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present invention with unnecessary detail.

[0047] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0048] It will also be understood that the term "and / or" used in the present description and appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0049] As used in the present specification and the appended claims, the term "if" may be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" may be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]," depending on the context.

[0050] In addition, in the description of the present specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0051] References to "one embodiment" or "some embodiments" in the present specification mean that a particular feature, structure, or characteristic described in conjunction with that embodiment is included in one or more embodiments of the present invention. Thus, phrases such as "in one embodiment," "in some embodiments," "in other embodiments," and "in yet other embodiments" appearing in various places in this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "including," "comprising," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0052] It should be understood that the order of execution of the steps in the following embodiments does not necessarily mean the order in which they are executed. The order in which each process is executed should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0053] In order to illustrate the technical solution of the present invention, a specific embodiment is used for illustration below. The 3D printing method of this embodiment is illustrated by taking the FDM process as an example, but is not limited to the FDM process, and can also be used for other 3D printing processes such as WAAM. A multi-degree-of-freedom robotic arm is used as an example for illustration, but is not limited to a multi-degree-of-freedom robotic arm structure, and can be other multi-degree-of-freedom structural devices; the 3D vision module in the method, also known as a 3D camera, is used to perform three-dimensional scanning of the object on the printing platform to obtain the three-dimensional spatial information of the object. The most direct result is the three-dimensional point cloud data of the object on the printing platform; the FDM print head module and the 3D vision module are installed at the end position of the robotic arm.

[0054] See also Figure 1 , is a flexible 3D printing method based on 3D vision provided in an embodiment of the present invention, which may include the following steps:

[0055] Step S100: obtaining a 3D model file of the target object, performing topological reconstruction on the 3D model file to obtain a patch model of the target object, and calculating a voxel model and a point cloud model of the 3D model based on the patch model.

[0056] First, read the three-dimensional model file of the target object to be printed. The three-dimensional model file records the three-dimensional model data of the target object, such as an STL file. The STL file is composed of the definitions of multiple triangular facets. The definition of each triangular facet includes the three-dimensional coordinates of each fixed point of the triangle and the normal vector of the triangular facet; then perform topological reconstruction on the three-dimensional model file to obtain the facet model G of the target object. Based on the facet model G, the voxel model Mv of the three-dimensional model and the point cloud model Mp of the three-dimensional model can be calculated.

[0057] Those skilled in the art know that any method in the prior art for performing topological reconstruction on a three-dimensional model file to obtain a mesh model falls within the scope of protection of the present invention; any method in the prior art for obtaining a voxel model and a point cloud model based on a mesh model falls within the scope of protection of the present invention.

[0058] Optionally, the volume of the sub-voxel model at the bottom of the target object targeted by the 3D vision-based flexible 3D printing method of the present invention is the largest.

[0059] Step S200: obtaining a three-dimensional size ratio R of the axis-aligned bounding box of the voxel model, and generating a bottom-largest filling voxel unit a on the bottom surface of the voxel model according to the three-dimensional size ratio R.

[0060] See also Figure 2 According to the existing axis-aligned bounding box calculation method, the axis-aligned bounding box E corresponding to the voxel model Mv is calculated, and the three-dimensional size ratio R of the axis-aligned bounding box E can be further calculated.

[0061] Calculate the voxel pattern shape of the bottom layer of the voxel model Mv, then select the voxel point located at the center of the pattern shape as the seed voxel a0, and record the current position.

[0062] According to the three-dimensional size ratio R calculated in the above steps, the seed voxel a0 is expanded in equal proportion until the expanded voxel contacts the outer voxels of the voxel model Mv, and the expansion of the seed voxel a0 is stopped.

[0063] Then, perform a tentative move in the horizontal direction of the XOY plane, as follows:

[0064] 1) Move the currently expanded voxel in the four directions of X+1, X-1 on the X axis and Y+1, Y-1 on the Y axis in sequence;

[0065] 2) Determine whether the expanded voxel after movement intersects with the outermost voxel of the voxel model Mv; if the expanded voxel units after four movements all intersect, it means that the current expanded voxel is the largest bottom segmentation voxel unit of the voxel model Mv, and the process ends; if the moved voxel unit does not intersect with the outermost voxel of the voxel model Mv, move the internal expanded voxel to that position, and then continue to expand proportionally according to the above voxel steps until the expanded voxel units after four movements all intersect, and obtain the largest filling voxel unit a at the bottom.

[0066] Step S300: segmenting the voxel model according to the filled voxel unit a to generate a preset number of sub-voxel units.

[0067] See also Figure 3 , the five planes of the largest voxel unit a in the voxel model Mv, namely the planes where the right splitting plane 1, the back splitting plane 2, the front splitting plane 3, the left splitting plane 4 and the upper splitting plane 5 are located, are used as splitting planes to split the voxel model Mv into 18 sub-voxel units.

[0068] Step S400: Decompose each subvoxel unit until there is a subvoxel whose bounding box size is smaller than the preset spatial size among the decomposed subvoxels, number all the subvoxels to obtain a model decomposition sequence tree, which includes several subvoxel sequences.

[0069] In this embodiment, the voxel model Mv of the three-dimensional model to be printed is used as the processing object, and a stack set S for constructing a voxel decomposition sequence tree and a sub-voxel decomposition result set R are established, and S and R are initially empty.

[0070] In the above step S300 , the central voxel located in the first layer is the voxel unit a, and the remaining 17 sub-voxels are pushed into the stack set S.

[0071] Traverse each subvoxel unit in S, calculate the three-dimensional size ratio Q of the axis-aligned bounding box of the subvoxel unit, and generate the largest filling voxel unit b on the bottom surface of the subvoxel unit according to the three-dimensional size ratio Q. The filling voxel unit b is used as the final subvoxel.

[0072] According to the filled voxel unit b, the subvoxel unit is recursively decomposed to generate several final subvoxels. The detailed steps are as follows:

[0073] Traverse each subvoxel unit according to the depth-first strategy and perform recursive decomposition, take out the subvoxel unit at the head of the stack in the stack set S in turn, and continue to decompose the subvoxel unit b according to the method in the above steps S200 and S300 until the stack stack S becomes empty again. The current given voxel model Mv is decomposed and the decomposed subvoxel result set R is obtained, which includes all the final subvoxels.

[0074] In this embodiment, the decomposition termination constraint of the voxel model Mv is defined: the axis-aligned bounding box corresponding to the decomposed voxel is calculated. If any dimensional component (X, Y or Z) of the axis-aligned bounding box is less than the given decomposition minimum value, it means that the voxel will no longer be decomposed, otherwise the voxel needs to be further decomposed; if any dimensional component (X, Y or Z) of the axis-aligned bounding box of the filling voxel unit a generated by the sub-voxel unit after the first segmentation is less than the given decomposition minimum value, the sub-voxel unit will not be decomposed.

[0075] Number all final subvoxels as follows:

[0076] Take the result set R as the processing object, calculate the axis-aligned bounding box of each final sub-voxel in the structure set R, traverse the maximum value (i.e., the first value) of the top surface of all final sub-voxels on the Z axis and the minimum value (i.e., the second value) of the bottom surface on the Z axis, and sort the final sub-voxels in R in ascending order according to the minimum value, and number the final sub-voxels according to the following rules: See Figure 4 , the rules are as follows:

[0077] 1) If the minimum values ​​of the final sub-voxels are the same, these final sub-voxels are assigned different letters for numbering, such as A, B, C...;

[0078] 2) If the maximum Z-axis value of the numbered final subvoxel is equal to the minimum Z-axis value of one or more unnumbered final subvoxels, then there is a Z-dependent relationship between the numbered final subvoxel and the unnumbered final subvoxel, where the lower surface of the latter is adjacent to the upper surface of the former. The numbering method of the latter is to use the numbering content of the former and append the index number of the corresponding subvoxel unit adjacent to its upper surface, such as: 1, 2, 3, etc.;

[0079] 3) If there are unnumbered sub-voxels whose minimum value is not on the surface of the forming base and whose minimum value is not equal to the maximum value of the numbered final sub-voxel, this situation is not within the scope of the present invention and can be avoided by adding a support structure to the three-dimensional model in the early stage.

[0080] The process of numbering the final sub-voxels in the result set R is the process of constructing the decomposition sequence tree T corresponding to the voxel model Mv.

[0081] For example, see Figure 4 , the bottom voxels are A1, B1, C1, and D1. Taking A1 as an example, the decomposition sub-voxel in contact with the upper surface of A1 is A11. There are two decomposition sub-voxel units in contact with the upper surface of A11, which are numbered A111 and A112 respectively. Continue to number the decomposition sub-voxel units in contact with the upper surface of A112, and they are numbered A1121 and A1122 respectively. Repeat the above steps until the numbering process of all decomposition sub-voxel units in the decomposition sequence tree T is completed, forming a series of ordered sub-voxel decomposition sequence trees T. The content of T is as follows:

[0082]

[0083] Step S500: Printing the target object according to the model decomposition sequence tree and the point cloud model to obtain a model of the target object.

[0084] A rough model of the target object is obtained according to the model decomposition sequence tree, and the internal filling area of ​​the rough model is printed using a first printing method.

[0085] Based on the model decomposition tree sequence T obtained in the above steps, the coarse model C composed of the sub-voxel set in T is located within the patch model G. The space within the coarse model C belongs to the internal filling area M1. The spatial structure of the internal filling area M1 is simple and does not contain the detailed features of the 3D model. A larger preset printing thickness f can be used for fast filling and printing. The specific printing process is as follows:

[0086] Starting from the internal filling area M1, according to the sub-voxel sequence in the model decomposition voxel sequence T, traverse from bottom to top, starting from the first layer, and print layer by layer according to the preset printing thickness f; the height h = i*f of the current layer i, calculate the slice contour where the slice position intersects with the rough model in T, use the conventional filling path for filling, and then use the robotic arm to drive the print head at the end to print according to the filling path.

[0087] A spatial region between the rough model and the surface of the patch model is obtained, and the spatial region is printed using a second printing method.

[0088] The spatial region M2 outside the rough model C and within the patch model G is close to the surface of the 3D model skin. This part of the spatial structure contains the surface details of the 3D model. A smaller preset printing thickness d is used for fine printing. The value of f is generally an integer multiple of d. Let f = k*d, where k is the number of prints per layer. The specific printing process is as follows:

[0089] The 3D vision module is used to scan and collect the point cloud model Mc of the printed entity on the current printing platform. By comparing the point cloud model Mp of the target object with the current point cloud model Mc, the remaining area R of the current layer i is obtained. This area belongs to the spatial area M2, and a layer thickness of d is used for fine printing.

[0090] When printing the remaining area R, the 3D vision module is used to position the printer nozzle near the inner edge of the remaining area R, and then the remaining area R is gradually covered starting from the current position according to the "Flood Fill Algorithm". At the same time, the 3D vision module needs to be used to continuously repair the remaining area R during the printing process to ensure that the remaining area R in the current layer can be accurately printed through 3D vision.

[0091] The remaining area R is printed k times until the layer thickness of the remaining area R reaches f, that is, d=f / k, and then all the remaining areas in the current i layer can be printed.

[0092] The above printing steps are performed on the remaining areas R of all layers until the height h=i*f of the current printing layer reaches the highest point position of the internal filling area M1. Then the spatial area of ​​the internal filling area M1 is printed, and the spatial area below the 3D model slice position h is also printed.

[0093] For the spatial region M2, that is, the part above the height of the rough model, the printing layer thickness of this region is selected as d, and the printing process of this region is as follows:

[0094] Calculate the height of the rough model after printing h1=i max*f, based on the height h1 and the point cloud model of the target object, calculate the area M2 to be printed above the height h1; where i max is the maximum number of layers corresponding to the rough model when f is the printing thickness of each layer, i max *f≤p<(i max +1)*f, p is the height of the rough model;

[0095] The 3D vision module is used to collect the point cloud model of the printed model on the printing platform to obtain the point cloud model Mc' of the current printed model. By comparing the point cloud model Mp of the target object with the point cloud model Mc' of the current printed model, the remaining area set U that needs to be printed is obtained.

[0096] Optionally, connected areas are regarded as the same area, and disconnected areas are regarded as different areas.

[0097] Calculate the current printing height h2 = h1 + j * d, where j is the number of times the area to be printed is printed, and traverse the positions of different areas in the set U. The different areas in the set U are areas that are not connected to each other. Use the flood filling algorithm to print each area with a height of h2 in the set U until the current printing height h2 reaches the maximum height of the point cloud model of the target object. At the same time, use the 3D vision module to continuously scan and collect the remaining areas during the printing process, and compare them with the original point cloud model Mp to ensure accurate printing at the edge positions.

[0098] For example, Figure 4 As shown, the tallest of the two ears of the cat model to be printed is the maximum height of the point cloud model of the target object. The left ear and the right ear are different areas in the set U. The height of the left ear is smaller than that of the right ear. When printing these two areas, the end condition for printing the left ear is that the left ear is printed to its maximum height, and the end condition for printing the right ear is that the right ear is printed to the maximum height of the point cloud model.

[0099] Repeat the above printing steps of the area to be printed until the printing height h2 reaches the maximum height of the point cloud model Mp of the target object, and the process can be terminated; at this point, the part of the spatial area M2 above the slice height h1 is printed, and the printed model of the target object is obtained.

[0100] The flexible 3D printing method based on 3D vision in this embodiment has the following characteristics:

[0101] (1) During the execution process, the strict and rigid dependency between the subsequent forming and manufacturing process and the previous model discrete process in the current 3D printing forming process is weakened. During the 3D printing forming process, the point cloud model of the printed entity is collected by 3D vision and is continuously compared and corrected with the original three-dimensional model, so that the three-dimensional entity in the actual printing can gradually approach the input three-dimensional model space geometric size. Finally, the printed three-dimensional model continues to "grow" from bottom to top and from inside to outside until the printing is completed; there is no need for the precise slicing of the three-dimensional model and path planning in the early stage, and even if the printing is interrupted due to uncontrollable abnormalities during the forming process, the flexible 3D printing forming method can continue to work, thereby improving the stability and reliability of the printing forming process;

[0102] (2) The 3D model decomposition sequence tree construction method can complete the coarse-grained orderly decomposition of the input 3D model, ensuring that the subsequent flexible 3D printing process is printed in an orderly manner from bottom to top and from inside to outside;

[0103] (3) A mixed layer thickness 3D printing method is proposed, in which the inner area is quickly printed with a large layer thickness, and the outer area is finely printed with a small layer thickness.

[0104] The present invention also provides a printing device, which includes a processor, a memory, and a printing device program stored in the memory and runnable on the processor. When the processor executes the printing device program, it implements the 3D vision-based flexible 3D printing forming method as described in the above embodiment.

[0105] The printing device also includes a 3D vision module, which is used to scan the target object currently being printed during the printing process to generate a point cloud model.

[0106] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0107] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0108] In the embodiments provided herein, it should be understood that the disclosed devices / clients and methods can be implemented in other ways. For example, the device / client embodiments described above are merely illustrative. For example, the division into modules or units is merely a logical functional division. In actual implementation, other division methods may be used, such as combining or integrating multiple units or components into another system, or omitting or not implementing certain features. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interface, or the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0109] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0110] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. A flexible 3D printing method based on 3D vision, characterized in that: include: Step S100: obtaining a three-dimensional model file of a target object, and performing topological reconstruction on the three-dimensional model file to obtain a surface model of the target object; Calculating a voxel model and a point cloud model of the target object according to the patch model; Step S200: obtaining a three-dimensional size ratio R of the axis-aligned bounding box of the voxel model, and generating a bottom-largest filling voxel unit a on the bottom surface of the voxel model according to the three-dimensional size ratio R; Step S300: segmenting the voxel model according to the filled voxel unit a to generate a preset number of sub-voxel units; Step S400: Decomposing each of the subvoxel units until a subvoxel with a bounding box size smaller than a preset spatial size exists among the decomposed subvoxels, numbering all the subvoxels to obtain a model decomposition sequence tree, wherein the model decomposition sequence tree includes a plurality of subvoxel sequences; Step S500: Printing the target object according to the model decomposition sequence tree and the point cloud model to obtain a model of the target object, including: Decomposing the model sequence tree, obtaining a rough model of the target object, and printing the inner filling area of ​​the rough model layer by layer from bottom to top, setting the printing thickness of each layer to f; Calculating the print height h=i*f of the i-th layer of the rough model, obtaining a slice contour where the slice position of the i-th layer intersects with the rough model of the target object based on the print height of the i-th layer, and performing fill printing on the i-th layer based on the slice contour; Obtain a current point cloud model of the i-th layer of the rough model, compare the current point cloud model with the point cloud model of the target object to obtain a remaining area of ​​the i-th layer, and print the remaining area of ​​the i-th layer according to a preset printing thickness d, where d=f / k, and k is an integer greater than 1.

2. The flexible 3D printing method based on 3D vision according to claim 1, characterized in that: Generating a bottom-largest filling voxel unit a on the bottom surface of the voxel model according to the three-dimensional size ratio R includes: According to the voxel pattern shape of the bottom layer of the voxel model, obtaining a voxel point located at the center of the voxel pattern shape of the bottom layer of the voxel model as a seed voxel; According to the three-dimensional size ratio R, the seed voxel is proportionally expanded until the expanded voxel contacts the outermost voxel of the voxel model; The expanded voxel is moved one voxel point in any of the four directions of the horizontal plane. If the expanded voxel does not intersect with the outermost voxels of the voxel model, the expanded voxel is further expanded according to the three-dimensional size ratio R until the expanded voxel intersects with the outermost voxels of the voxel model by moving one voxel point in any of the four directions of the horizontal plane.

3. The flexible 3D printing method based on 3D vision according to claim 1, characterized in that: According to the filled voxel unit a, the voxel model is segmented to generate a preset number of sub-voxel units, including: The plane where the remaining five surfaces of the filling voxel unit a excluding the bottom surface are located is used as a segmentation plane, and the voxel model is divided into eighteen sub-voxel units, and the filling voxel unit a is used as the final sub-voxel.

4. The flexible 3D printing method based on 3D vision according to claim 3, characterized in that: Decomposing each of the subvoxel units includes: For any subvoxel unit m excluding the filling voxel unit a, obtaining a three-dimensional size ratio Q of the axis-aligned bounding box of the subvoxel unit m, generating a filling voxel unit b with the largest bottom on the bottom surface of the subvoxel unit m according to the three-dimensional size ratio Q, and using the filling voxel unit b as the final subvoxel; The subvoxel unit m is recursively decomposed to generate several final subvoxels.

5. The flexible 3D printing method based on 3D vision according to claim 4, characterized in that: All subvoxels are numbered to obtain the model decomposition sequence tree, including: Obtain a first value of the top surface of each final sub-voxel on the Z axis and a second value of the bottom surface on the Z axis; Taking each final sub-voxel with the smallest second value as the bottom sub-voxel, and numbering each bottom sub-voxel; For any underlying subvoxel l , compare the second value of the unnumbered bottom sub-voxel with the bottom sub-voxel l The first value of the same underlying subvoxel is the same as the underlying subvoxel l Perform associated numbering to obtain the underlying subvoxels l The corresponding subvoxel sequence; The model decomposition sequence tree is obtained according to the sub-voxel sequence corresponding to each underlying sub-voxel.

6. The flexible 3D printing method based on 3D vision according to claim 1, characterized in that: Printing the remaining area of ​​the i-th layer according to a preset printing thickness d includes: The nozzle of the printer is positioned at the inner edge of the remaining area through the 3D vision module, and the remaining area of ​​the i-th layer is printed k times according to the flood filling algorithm until the thickness of the remaining area of ​​the i-th layer is equal to the printing thickness f.

7. The flexible 3D printing method based on 3D vision according to claim 1, characterized in that: The method further comprises: Calculate the height h1=i of the rough model after printing max *f, based on the height h1 and the point cloud model of the target object, calculate the area to be printed above the height h1; i max is the maximum number of layers corresponding to the rough model when f is the printing thickness of each layer, i max *f≤p<(i max +1)*f, p is the height of the rough model; The area to be printed is printed according to the preset printing thickness d to obtain a model of the target object.

8. The flexible 3D printing method based on 3D vision according to claim 7, characterized in that: Printing the to-be-printed area according to the preset printing thickness d includes: Acquire a point cloud model of a model with a current printing height of h1, and compare the point cloud model of the model with the current printing height of h1 with the point cloud model of the target object to obtain a set U of the to-be-printed area; Calculate the current printing height h2 = h1 + j * d, where j is the number of times the area to be printed is printed. Traverse the different area positions in the set U and use the flood filling algorithm to print each area with a height of h2 in the set U until the current printing height h2 reaches the maximum height of the point cloud model of the target object.

9. A printing device, characterized in that: The printing device includes a processor, a memory, and a printing device program stored in the memory and runnable on the processor. When the processor executes the printing device program, the 3D vision-based flexible 3D printing method according to any one of claims 1 to 8 is implemented.

Images