A method and system for rapid reconstruction and adaptive printing of a three-dimensional model

Abstract

The application provides a three-dimensional model rapid reconstruction and adaptive printing method and system, the method comprising: obtaining object acquisition data of a target object, establishing a three-dimensional reconstruction model, and limiting a local allowed geometry range; slicing a current surface and two boundary states along a printing direction, comparing three local entity cross-sections layer by layer in the same printing layer, and forming a manufacturing constraint model; generating layer-by-layer slicing data according to the same printing direction and a basic layer thickness, mapping manufacturing influences to corresponding printing layers, taking a continuous printing layer position as a main section, combining the local allowed geometry range to determine an entry section and an exit section, forming manufacturing events and associating printing parameter categories, and constructing a printing task; calling normal parameters in ordinary layers and calling corresponding adaptive parameters in manufacturing event layer sections during printing, and obtaining a three-dimensional printing entity. The application realizes the unity between the rapid acquisition of a three-dimensional model and an actual adaptive printing process.

Description

Technical Field

[0001] This invention belongs to the field of 3D printing, and in particular relates to a method and system for rapid reconstruction and adaptive printing of 3D models. Background Technology

[0002] With the development of 3D scanning, photogrammetry, depth sensing, and additive manufacturing technologies, the rapid acquisition of 3D digital models from photographs, scanned data, or sketches of real objects, followed by physical manufacturing, has been increasingly applied to personalized toys, cultural and creative products, ornament replication, reverse engineering of parts, and rapid repair. Current technologies typically involve first acquiring the shape information of the target object using mobile terminal cameras, industrial cameras, depth cameras, structured light scanning equipment, or laser scanning equipment. Then, a complete 3D model is formed using multi-view matching, point cloud registration, surface generation, mesh reconstruction, and hole repair. Subsequently, slices are generated according to a predetermined printing direction and layer thickness, and the 3D printing equipment completes layer-by-layer manufacturing based on pre-set forming parameters. In photopolymer 3D printing scenarios, it is also necessary to set process parameters such as exposure, curing, and interlayer separation based on different cross-sectional structures. To shorten 3D model generation time, existing rapid reconstruction methods typically reduce the number of acquisition views, lower sampling density, or directly perform surface completion on occluded and missing areas. Their primary goal is to quickly obtain a complete, continuous, and slicable model. To improve printing stability, existing printing methods usually set slicing and process parameters based on the already determined model geometry, or adjust subsequent printing processes based on abnormal conditions encountered during printing. Both of these methods can meet general manufacturing requirements when data is sufficient and model geometry can be reliably determined. However, for applications where users rapidly generate 3D models of customized toys or other products from limited photographs, partial scans, or sketches, a significant disconnect remains between reconstruction and printing. Due to factors such as obstruction, insufficient observation direction, missing texture, reflection, and scanning blind spots, areas such as the connection between the arm and torso, the base of the ear, the fingertips, the interior of grooves, and shell openings often cannot be uniquely determined by the acquired data. If supplementary data acquisition and repeated fine reconstruction are carried out until all areas achieve uniform accuracy, the waiting time from acquisition to printing will be significantly extended. If the completed and determined model is used directly, although the model is visually complete, the subsequent printing system cannot distinguish which surfaces are fully determined by actual observation and which surfaces still have reasonable geometric variation space. More importantly, there is no simple correspondence between the magnitude of local geometric errors and their actual manufacturing impact. When the surface position of a large continuous curved surface changes to a certain extent, the connectivity and forming state of the layer-by-layer cross-sections may remain stable. However, when the position of the limb connection or small structure changes slightly, it may cause the entities in adjacent printed layers to change from connection to separation, cause a certain local cross-section to appear prematurely or disappear late, cause the fine connection to change from stable forming to difficult to form, or cause the originally open internal space to form a closed structure. For photopolymer printing, these changes will further change the local curing area and the interlayer separation process.Existing reconstruction methods typically represent these differences as geometric errors or completion results. Furthermore, existing printing methods generally treat the input model as a predetermined final geometry. Therefore, local geometric uncertainties generated during the reconstruction phase are difficult to incorporate into subsequent slicing and printing tasks after the model is completed. Printing equipment also struggles to pre-identify manufacturing-sensitive layers caused by the rapid reconstruction process itself. This results in a situation where, on the one hand, a large number of areas that will not actually change the manufacturing state are meticulously processed to reduce reconstruction errors; on the other hand, local areas that may cause changes in connectivity, fine structures, closure states, or curing ranges can only be processed according to the standard, predetermined model during the printing phase.

[0003] Therefore, how to maintain rapid reconstruction efficiency while preserving the actual range of geometric changes in under-observed areas, and further determine whether these geometric changes will cross the actual layer-by-layer manufacturing state, and then transform this manufacturing influence into hierarchical information that can directly participate in the execution of the printing equipment, is a practical engineering problem that needs to be solved to achieve effective connection between rapid reconstruction of 3D models and subsequent adaptive printing. Summary of the Invention

[0004] The purpose of this invention is to propose a method and system for rapid reconstruction and adaptive printing of three-dimensional models, thereby solving the above-mentioned problems.

[0005] To achieve the above objectives, a method for rapid reconstruction and adaptive printing of three-dimensional models is provided in a first aspect of the present invention, the method comprising the following steps: S1. Acquire object acquisition data of the target object to establish the current three-dimensional surface, including determining the current surface of the three-dimensional surface and the insufficient observation area. For the current surface of the insufficient observation area, determine the local allowable geometric range based on the current surface to generate a three-dimensional reconstruction model. The three-dimensional reconstruction model includes determining the current surface of the three-dimensional surface and the insufficient observation area, as well as the corresponding local allowable geometric range. The current surface of the insufficient observation area is determined based on the region where there are local geometric changes. S2. Based on the aforementioned 3D reconstruction model, determine three corresponding local solid sections along the printing direction that correspond to the current 3D surface and the local allowable geometry within the same printing layer, wherein the 3D surface portion is kept consistent; compare the layer-by-layer manufacturing state among the three local solid sections within the same local allowable geometry. When any geometric state undergoes a manufacturing state change on any affected printing layer, the location corresponding to the local allowable geometric range is marked as the manufacturing influence area, and the manufacturing influence type and the location of the continuous printing layer are recorded. At the same time, it is associated with the three-dimensional reconstruction model to form a manufacturing constraint model; the geometric state corresponding to the inner allowable boundary and the geometric state corresponding to the outer allowable boundary within the local allowable geometric range. S3. Based on the manufacturing constraint model, generate layer-by-layer slicing data according to the same printing direction and base layer thickness as in S2, map the manufacturing influence area to the corresponding printing layer, take the continuous printing layer position as the main layer segment, combine the local allowable geometric range to determine the entry segment and exit segment, form manufacturing events and associate them with printing parameter categories to obtain manufacturing event information, so as to construct a printing task; the printing task includes layer-by-layer slicing data and manufacturing event information. S4. Read the printing task and determine the corresponding equipment adaptation parameter combination based on the manufacturing event information; call the regular parameters in the ordinary printing layer, and call the equipment adaptation parameter combination in the entry segment, main segment and exit segment of the corresponding manufacturing event to print layer by layer to obtain the printing results of each layer; after completing all printing layers, summarize to obtain the three-dimensional printed entity.

[0006] Furthermore, the acquisition of object collection data of the target object includes: When using multi-view photographs, keep the target object still, acquire images from multiple viewing directions, and ensure that there are repeatable observation areas of the object surface between adjacent viewpoints; A correspondence is established based on the common visible object features in adjacent images. The corresponding features are used to obtain the position and direction of each shooting perspective relative to the coordinate relationship of the same object, so that the spatial positions obtained from different perspectives are unified into the same three-dimensional reconstructed coordinate relationship.

[0007] Furthermore, determining the local permissible geometry based on the current surface includes: Extract the established 3D surface boundaries around the currently under-observed region, and generate a continuous surface towards the missing region along the extension direction of the original surface at the boundary. The position of the surface to be verified is moved step by step along the inner and outer directions of the current local surface, and the object acquisition data is used to determine whether the position still matches the actual observation. When two adjacent attempts correspond to satisfying and exceeding the existing observation constraints, respectively, the search range is further narrowed between the two until the local boundary position is stable, thus forming the inner and outer allowable boundaries of the current local region.

[0008] Furthermore, the comparison of layer-by-layer manufacturing states includes determining the entity connectivity state: The interior of the closed contour obtained by slicing is taken as the solid region, and the same connectivity label is assigned to regions that are in contact with each other and can be connected by continuous solid paths. When a local structure in the current section is connected to the main body, but the two are separated in the section corresponding to the inner allowable boundary or the section corresponding to the outer allowable boundary, the local position is recorded as the manufacturing influence area where the connection state changes. When a local structure in the current section separates from the main body, and a new solid connection is formed between the section corresponding to the inner allowable boundary or the section corresponding to the outer allowable boundary, it is also recorded as a manufacturing influence area where the connection state changes.

[0009] Furthermore, the comparison of layer-by-layer manufacturing states includes determining whether a local entity appears or disappears: For local structures extending outward from the main body, check layer by layer whether the corresponding entity regions in the current section, the section corresponding to the inner allowable boundary, and the section corresponding to the outer allowable boundary exist simultaneously; When a local solid has been formed on a certain 3D surface in a certain printing layer, and the geometric state corresponding to the inner allowable boundary or the outer allowable boundary has not yet intersected with the printing layer, it is recorded as a manufacturing influence area related to the cross-section existence state.

[0010] Furthermore, the determination of the entry and exit segments based on the locally permissible geometric range includes: The number of additional printing layers required is determined based on the maximum positional variation of the local permissible geometric range along the printing direction; Add corresponding entry and exit segments before and after the main layer segment, and check layer by layer whether the local structure corresponding to the manufacturing impact area still exists in the newly added printing layer; When a newly added printing layer does not have a corresponding local structure, the manufacturing event in the current direction terminates at the previous printing layer where the local structure still exists.

[0011] Furthermore, the formation of the manufacturing event includes: The local locations of the same manufacturing influence area within consecutive printing layers are organized into manufacturing events according to their spatial continuity. When the local action positions in two adjacent layers correspond to the same defined three-dimensional surface structure, and the local action position of the upper layer extends along the printing direction and continuously overlaps with the lower layer, the two adjacent layers are classified into the same manufacturing event. The manufacturing impact type remains the established type, and the local scope of each layer is preserved according to the printing layer order.

[0012] Furthermore, the associated printing parameter categories include: The printing parameter categories are determined based on the type of manufacturing impact; among them, changes in connection state correspond to the connection structure adaptation parameter category, changes in fine connection forming correspond to the fine structure adaptation parameter category, the appearance or disappearance of cross-sections correspond to the local new structure adaptation parameter category, changes in open and closed states correspond to the closed structure adaptation parameter category, and changes in curing range correspond to the cross-section change adaptation parameter category.

[0013] Furthermore, the step of calling the device adaptation parameter combination for layer-by-layer printing in the entry, main, and exit segments of the corresponding manufacturing event includes: When multiple manufacturing events exist in the same printing layer, the entity region corresponding to each event is determined according to the local action position recorded in the printing task record; For printing devices that support local exposure control, each local area calls the exposure configuration corresponding to its own manufacturing effect type, while the remaining areas execute the normal exposure configuration of that layer. For parameters that need to be executed uniformly across the entire layer, the uniform motion configuration of this layer is determined according to the predetermined priority order of manufacturing states. The priority order is as follows: connection state change, open and closed state change, cross-section appearance or disappearance, fine connection forming change, and curing range change. Specifically, the device adaptation parameter combination is invoked at the first layer of the entry segment of the manufacturing event and continues until the main segment and the exit segment; after the last layer of the exit segment is completed, the normal printing parameters are restored from the next normal printing layer.

[0014] A second aspect of the present invention provides a system for rapid reconstruction and adaptive printing of three-dimensional models, the system comprising: A 3D reconstruction processing unit is used to acquire object acquisition data of the target object to establish the current 3D surface, including determining the current surface of the 3D surface and the current surface of the under-observed region. For the current surface of the under-observed region, the local allowable geometric range is determined based on the current surface to generate a 3D reconstruction model. The 3D reconstruction model includes determining the 3D surface, the current surface of the under-observed region, and the corresponding local allowable geometric range. The current surface of the under-observed region is determined based on the region where there are local geometric changes. The manufacturing impact analysis unit is used to determine, based on the 3D reconstruction model, three corresponding local solid sections of the current 3D surface and the local allowable geometry within the same printing layer along the printing direction, wherein the 3D surface portion is determined to be consistent; and to compare the layer-by-layer manufacturing state among the three local solid sections within the same local allowable geometry. When any geometric state undergoes a manufacturing state change on any affected printing layer, the location corresponding to the local allowable geometric range is marked as the manufacturing influence area, and the manufacturing influence type and the location of the continuous printing layer are recorded. At the same time, it is associated with the three-dimensional reconstruction model to form a manufacturing constraint model; the geometric state corresponding to the inner allowable boundary and the geometric state corresponding to the outer allowable boundary within the local allowable geometric range. The printing task generation unit is used to generate layer-by-layer slicing data according to the manufacturing constraint model, with the same printing direction and base layer thickness as S2, map the manufacturing influence area to the corresponding printing layer, take the continuous printing layer position as the main layer segment, combine the local allowable geometric range to determine the entry segment and exit segment, form a manufacturing event and associate it with the printing parameter category to obtain manufacturing event information, so as to construct a printing task; the printing task includes layer-by-layer slicing data and manufacturing event information. The printing control unit is used to read the printing task, determine the corresponding equipment adaptation parameter combination based on the manufacturing event information, call the conventional parameters in the normal printing layer, and call the equipment adaptation parameter combination in the entry segment, main segment and exit segment of the corresponding manufacturing event to print layer by layer to obtain the printing results of each layer; after completing all printing layers, the results are summarized to obtain the three-dimensional printed entity.

[0015] The beneficial technical effects of the present invention are at least as follows: This invention addresses the contradiction that local geometry is not fully determined in rapid 3D reconstruction while subsequent printing requires the determination of manufacturing behavior. It continuously transmits the geometric uncertainty information in the 3D reconstruction results to the actual printing process, so that reconstruction and printing are continuously associated around the same manufacturing state.

[0016] First, when forming a complete 3D reconstruction model based on the collected object data, the 3D surface is determined for areas with sufficient observation data. For areas affected by occlusion, insufficient viewpoint, or missing sampling, while forming the current surface that can be used for slicing, the local permissible geometric range limited by the actual observation conditions is retained. This allows the rapid reconstruction result to simultaneously express the current printable shape and the real geometric changes that may still occur in areas with insufficient observation. Subsequently, combined with the predetermined printing direction, base layer thickness, and the stable forming conditions of the target photopolymerization 3D printing equipment for the current material, the cross-sectional state of the current 3D surface and the local permissible geometric range in the same printing layer is compared. This is done based on whether the local geometric changes cause changes in the solid connection relationship, the appearance or disappearance of local cross-sections, changes in the forming state of fine connections, changes in the internal open and closed state, or changes in the curing range. Significant changes occur, identifying the local areas with real manufacturing impact, and maintaining their correspondence with the original 3D geometry to form a manufacturing constraint model. These spatial manufacturing impacts are then converted into specific manufacturing events according to the actual printing sequence. The event's main scope is formed based on the continuous printing layers where the manufacturing impact is located, and the event's entry and exit ranges are determined by combining the possible positional changes of the local allowable geometric range along the printing direction. This ensures that each manufacturing event corresponds to the specific object structure, the actual slice position, and the corresponding manufacturing impact type. Finally, the photopolymer 3D printing equipment directly reads the formed printing task. For ordinary printing layers, the conventional printing configuration of the current material is used; for the layers where manufacturing events are located, the pre-calibrated adaptive printing configuration is called according to the corresponding manufacturing impact type, and this process is continuously executed along with the actual local position in each layer of slicing until the 3D solid manufacturing is completed.

[0017] Therefore, this invention gradually transforms the local geometrical changes in the rapid reconstruction phase into manufacturing impacts, hierarchical manufacturing events, and actual printing actions. This allows geometrical uncertainties that do not change the manufacturing state to directly enter subsequent processing while maintaining the current reconstruction results. Meanwhile, local uncertainties that may actually change the layer-by-layer forming state are addressed in the corresponding printing layers. This reduces the time consumption caused by uniform and detailed reconstruction of all under-observed areas, while improving the printing adaptability of the rapid reconstruction model in critical connection structures, small structures, closed structures, and areas with cross-sectional changes. This achieves unity between rapid acquisition of the 3D model and the actual adaptive printing process. Attached Figure Description

[0018] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0019] Figure 1 This is a flowchart of a method for rapid reconstruction and adaptive printing of a three-dimensional model according to an embodiment of the present invention.

[0020] Figure 2 This is a framework diagram of a three-dimensional model rapid reconstruction and adaptive printing system according to an embodiment of the present invention. Detailed Implementation

[0021] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0022] like Figure 1 As shown in the figure, an embodiment of the present invention provides a method for rapid reconstruction and adaptive printing of a 3D model, the method comprising: S1. Acquire object acquisition data of the target object to establish the current three-dimensional surface, including determining the current surface of the three-dimensional surface and the insufficient observation area. For the current surface of the insufficient observation area, determine the local allowable geometric range based on the current surface to generate a three-dimensional reconstruction model. The three-dimensional reconstruction model includes determining the current surface of the three-dimensional surface and the insufficient observation area, as well as the corresponding local allowable geometric range. The current surface of the insufficient observation area is determined based on the region where there are local geometric changes. S2. Based on the aforementioned 3D reconstruction model, determine three corresponding local solid sections along the printing direction that correspond to the current 3D surface and the local allowable geometry within the same printing layer, wherein the 3D surface portion is kept consistent; compare the layer-by-layer manufacturing state among the three local solid sections within the same local allowable geometry. When any geometric state undergoes a manufacturing state change on any affected printing layer, the location corresponding to the local allowable geometric range is marked as the manufacturing influence area, and the manufacturing influence type and the location of the continuous printing layer are recorded. At the same time, it is associated with the three-dimensional reconstruction model to form a manufacturing constraint model; the geometric state corresponding to the inner allowable boundary and the geometric state corresponding to the outer allowable boundary within the local allowable geometric range. S3. Based on the manufacturing constraint model, generate layer-by-layer slicing data according to the same printing direction and base layer thickness as in S2, map the manufacturing influence area to the corresponding printing layer, take the continuous printing layer position as the main layer segment, combine the local allowable geometric range to determine the entry segment and exit segment, form manufacturing events and associate them with printing parameter categories to obtain manufacturing event information, so as to construct a printing task; the printing task includes layer-by-layer slicing data and manufacturing event information. S4. Read the printing task and determine the corresponding equipment adaptation parameter combination based on the manufacturing event information; call the regular parameters in the ordinary printing layer, and call the equipment adaptation parameter combination in the entry segment, main segment and exit segment of the corresponding manufacturing event to print layer by layer to obtain the printing results of each layer; after completing all printing layers, summarize to obtain the three-dimensional printed entity.

[0023] In this embodiment, specifically, in step S1, the object acquisition data is obtained by a mobile terminal camera, industrial camera, depth camera, structured light scanning device, or laser scanning device. When a hand-drawn sketch is used as supplementary input, the sketch is used to express the object's outer contour, local structural positions, and the connection relationships between structures. When using multi-view photographs, the object to be printed is kept stationary, and images are acquired from the front, side, oblique side, and observation directions that can cover the main obstructed areas, ensuring that there are repeatable object surface areas between adjacent viewpoints. The camera's internal imaging parameters are preferentially read from the device's existing calibration results. When recalibration is required, calibration patterns with known geometric relationships are obtained by photographing them. The camera shooting position and observation direction corresponding to each image are then established based on the common visible object features in adjacent images, and these corresponding features are used to determine the position and direction of each shooting viewpoint relative to the same object's coordinate relationship, so that the spatial positions obtained from different viewpoints are unified into the same three-dimensional reconstructed coordinate relationship. When using depth cameras, structured light devices, or laser scanning devices, spatial sampling results of the object's surface are acquired from multiple viewing directions. The spatial alignment of the scan data is achieved by utilizing recurring object surfaces in adjacent scans, unifying data obtained from different acquisition directions into the same object's spatial relationship. Taking a customized toy as an example, images of the toy's front, left and right sides, left and right oblique sides, and back can be acquired separately, ensuring that the main structures such as the head, torso, and arms partially overlap in at least multiple adjacent viewpoints. If areas located between the arms and torso, on the inner side of the legs, or in local recesses are continuously occluded during acquisition, these locations are retained as under-observed areas in subsequent reconstruction.

[0024] In one specific embodiment, the current three-dimensional surface of the target object is established based on object acquisition data. When using multi-view photographs, contour inflection points, texture inflection points, structural edges, or stable visible local features that correspond to the same entity location are identified from different photographs. The three-dimensional positions of these locations in object space are determined by combining the viewing directions corresponding to each photograph. Then, a local three-dimensional surface is generated based on the spatial proximity and surface continuity relationships between adjacent three-dimensional positions. When using scan data, repeated surface positions obtained from different scanning directions are registered, and then connected according to adjacent spatial sampling positions and their local surface directions to form a three-dimensional surface.

[0025] During the formation of a local surface, the actual observation basis for that local surface is recorded simultaneously: when a local surface can be jointly defined by data from different observation directions, and the spatial positions obtained from these observations are consistent, the region is designated as a defined 3D surface; when a local location is only observed from one side, is continuously occluded, or has continuously missing scan data, the local location is marked as a region with local geometric changes; for regions with local geometric changes, the boundaries of the surrounding defined 3D surfaces are extracted, and a continuous surface is generated along the extension direction of the original surface at the boundary, so that the continuous surface maintains a smooth connection with the surrounding defined surfaces and enables the overall model to form a continuous closed structure; this continuous surface is used as the current surface of the region with insufficient observation and is associated with and saved with that region. The consistency of observations is judged directly using the allowable spatial position deviation under the calibration conditions of the acquisition equipment as the criterion. When the same local surface position obtained from different observation directions all fall within the allowable deviation range, it indicates that each observation forms a consistent constraint on the spatial position of the surface. Taking a toy head as an example, if the tip of the nose appears in both the frontal and oblique side views, the two different viewing directions can jointly define the spatial position of the tip of the nose; while if the back connecting surface of the arm close to the torso is always occluded, its current connecting surface is continuously completed by the surrounding determined structure, while continuing to retain the corresponding geometric change space.

[0026] For regions with local geometric changes, the permissible local geometric range is determined based on the current surface. First, the defined 3D surface boundaries surrounding the region are extracted, and continuous surfaces are generated towards the missing areas based on the extension direction of the original surfaces at the boundaries, forming a complete outer surface that can be read by subsequent printing slices. Then, the position of the surface to be verified is gradually moved along the inner and outer directions of the current local surface, and existing object acquisition data is used to determine whether the position still matches the actual observation. For photographic data, the moved local surface is reprojected onto all images where the region or its boundaries can be observed. The relationship between the reprojected object edges and the defined object contours in the photographs is compared, and the spatial relationship between the local surface and the surrounding defined 3D surfaces is determined. If the moved surface remains within the space defined by the existing object contours, maintains continuity with the surrounding defined structures, and does not enter the space occupied by other defined surfaces, this position continues to be included in the permissible local geometric range. For scanned data, the moved surface is compared with the already obtained defined scanned surfaces and the space where the scanning equipment can clearly exclude the existence of entities, retaining the surface position consistent with the defined scan results. When searching step by step in one direction, if two adjacent attempts correspond to "satisfying existing observation constraints" and "exceeding existing observation constraints" respectively, the search range is further narrowed between the two until the local boundary position can stably fall within the geometric range that the subsequent slices of the target printing device can distinguish, thereby forming the inner allowable boundary and the outer allowable boundary of the local region respectively.

[0027] In one specific embodiment, taking the dorsal connection area between a toy arm and torso as an example, the frontal image can determine the outer contour of the arm, the side image can determine the local thickness of the arm, and the oblique side image can determine a portion of the connection edge between the arm and torso, while the middle of the connection surface is continuously occluded. The system first uses the already determined arm surface, torso surface, and visible connection edge to form a continuous current connection surface, and then gradually adjusts its spatial position along the possible change direction of the connection surface. After a certain adjustment, if the outer contour of the arm in the existing photos remains consistent, the connection surface and the determined surface of the torso remain reasonably connected, and it does not penetrate into the solid area already determined by multi-view data, then the adjusted geometry is within the allowable range; if after further adjustment the reprojection result crosses the already observed arm contour edge, or the connection surface enters the interior of the determined torso surface, then the position that still satisfies the observation relationship in the previous step is taken as the boundary of the allowable range in that direction. For large, slightly occluded areas on the back of a toy, the permissible geometry is typically compressed by the visible contours on both sides, as the left and right oblique images together constrain most of the back's shape. For deeply occluded areas located between limbs and torso, the surrounding observable constraints are relatively fewer, and the permissible geometry is defined by adjacent defined surfaces and known connecting edges. This locally permissible geometry directly reflects the actual geometric changes in that area under the current acquisition conditions and can be used in subsequent steps to determine whether these changes alter the layer-by-layer printing state.

[0028] After processing each local region, the current surface of the determined 3D surface, the under-observed regions, and the corresponding local allowable geometric ranges are all written into the same 3D reconstruction model. The determined regions save the 3D surface constrained by the object acquisition data; the under-observed regions save the current surface that allows the overall model to be continuously closed, while also associating regions with local geometric changes defined by the inner and outer allowable boundaries. The model's continuity processing smoothly extends along the boundaries of the determined surfaces to the locally missing regions, forming a complete closed structure on the current 3D surface and maintaining the association between the formed local allowable geometric ranges and the corresponding surface regions. The final output is a 3D reconstruction model, which includes both the complete current 3D surface that can be directly processed in subsequent slicing steps and the local allowable geometric ranges corresponding to each under-observed region. This allows the next step to directly compare the impact of the current surface and other possible surface states within the allowable range on the layer-by-layer manufacturing results.

[0029] In this embodiment, specifically in step S2, a photopolymerization 3D printing device is used for manufacturing. Simultaneously, the device reads the pre-determined printing direction, base layer thickness, stable connection conditions, stable fine structure forming conditions, and allowable cross-sectional variation range for the current material. The stable connection conditions and stable fine structure forming conditions are obtained by printing standard structural specimens with different connection widths and fine structure dimensions using the same device and the current material. The process boundary corresponding to the specimen that can be continuously and completely formed is used as the judgment condition. The allowable cross-sectional variation range is obtained by printing transitional structural specimens with different degrees of cross-sectional variation between adjacent layers. The range corresponding to the specimen that can maintain continuous curing and normal interlayer separation is used as the judgment basis, and this range is saved together with the process configuration of the current material. The current 3D surface represents the geometric state actually prepared for printing according to the existing reconstruction results. The 3D surface is determined as a fixed reference in this step, and the local allowable geometric range represents the local geometric change boundary formed in step one based on the object acquisition data.

[0030] In one specific embodiment, for each local permissible geometric range, the continuous printing layer covering it is first determined along a predetermined printing direction. The current 3D surface, the geometric state corresponding to the inner permissible boundary, and the geometric state corresponding to the outer permissible boundary are intersected with the same printing layer to form three corresponding local solid cross-sections. The determined 3D surface portion in the three cross-sections remains consistent, and only the local geometry corresponding to the insufficiently observed area is changed. Taking the dorsal connection region between the toy arm and torso as an example, local cross-sections formed by the current connection surface, the connection surface corresponding to the inner permissible boundary, and the connection surface corresponding to the outer permissible boundary are obtained on the same printing layer. This allows for comparison of whether the same uncertain reconstruction region, within the permissible range of variation, will change the actual layer-by-layer manufacturing state.

[0031] For each local permissible geometry, first determine the continuous printing layers to be covered based on its lowest and highest positions in the printing direction, and then use these layers as the affected printing layers for that range.

[0032] For each affected printing layer, the connectivity state of the entities is first determined. Specifically, the interior of the closed contour obtained from slicing is considered the entity region. Regions that are in contact and can be connected by continuous entity paths are assigned the same connectivity marker. Then, it is checked whether the local structure still belongs to the same connected region as the main structure determined in step one. When a local structure in the current section is connected to the main body, but the two are separated in the section corresponding to the inner or outer allowable boundary, the corresponding local position is directly recorded as the manufacturing impact region where the connectivity state has changed. When the current state is separation, and a certain restriction state within the allowable range forms a new entity connection, it is also recorded as a change in connectivity state. For cases where all three sections maintain geometric connectivity, but the connection portion narrows significantly with allowable geometric changes, it is further determined whether this connection can be stably maintained in continuous printing layers. Here, the covering idea of ​​"the proportion of the common part to the reference set" in set theory and the persistence idea of ​​"the proportion of the number of times the condition is satisfied to the number of observations" in discrete statistics are used to determine the spatial retention degree and inter-layer persistence of the same connection structure. Since the two parts describe the section retention and continuous maintenance relationship of the same manufacturing phenomenon respectively, their proportions are calculated separately and then multiplied to obtain the connectivity persistence coefficient: ; in, This represents the connectivity persistence coefficient corresponding to the currently analyzed printed layer; The area of ​​the solid connection region that is jointly reserved at the same connection position by the current cross section, the cross section corresponding to the inner allowable boundary, and the cross section corresponding to the outer allowable boundary is obtained by finding the common part of the three sliced ​​solid regions. The area of ​​the connection region formed by the current 3D surface in this printing layer is obtained directly from the solid region connecting the local structure and the main body in the current slice; The number of printing layers selected around the current printing layer in a continuous analysis layer in which all three geometric states maintain the same connectivity relationship; This represents the total number of printed layers within this continuous analysis layer. and Using the same unit of area, and All are the number of printing layers, and only in Greater than zero The continuity coefficient is calculated when it is greater than zero. These are dimensionless proportional values. The first proportion in the formula reflects the degree to which the connected entities are still jointly preserved after allowing geometric changes, and the second proportion reflects the degree of continuity of this connection relationship in the layer-by-layer manufacturing direction. Both are derived from proportional relationships and therefore can be directly combined. The equipment pre-stores a stable connection lower limit for the current material. After printing several known connection structure specimens using the same equipment, the same material, and the same basic process, this lower limit is determined based on the connection continuity coefficient corresponding to the specimens that can be continuously and completely formed, and is saved along with the material and process configuration.

[0033] In one specific embodiment, for example, the area of ​​the connection region at the base of a toy's ear in the current analysis layer is 24 square millimeters, and the area of ​​the connection region jointly retained by the current cross-section, the inner allowable boundary corresponding cross-section, and the outer allowable boundary corresponding cross-section is 18 square millimeters. Five consecutive printed layers are selected for inspection, and four of these layers maintain the same connection relationship under all three geometric states. The current connection continuity coefficient is calculated by dividing 18 by 24, then multiplying by 4 and dividing by 5, resulting in 0.6. If the lower limit of stable connection in the current material configuration of the target device is 0.65, then although this location still maintains geometric connectivity in the current analysis layer, the allowable geometric range has caused the degree of spatial preservation of the connected entity and the degree of continuity layer maintenance to both fall below the stable connection condition. Therefore, the area corresponding to the ear base is recorded as a manufacturing influence area related to connection stability. If the calculation result at the same location is higher than the lower limit of stable connection, and the three restriction states maintain connection in all relevant layers, then the subsequent manufacturing state judgment continues. When the current cross-section itself has no connection region, it is directly recorded based on the change in connectivity. Therefore, this formula is only used when there is connection in all three geometric states but the degree of connection stability may change.

[0034] After determining the connection status, for areas that still maintain the same connectivity, further checks are made on whether local entities appear or disappear, whether small structures cross stable forming conditions, and whether the internal space undergoes an open-close transition. For structures extending outward from the main body, such as toy fingers, ear tips, tails, or decorative protrusions, the corresponding entity regions in the current cross-section and the two restricted cross-sections are checked layer by layer to see if they exist simultaneously. If a local entity has already formed on a certain printing layer on the current 3D surface, and the other restricted state within the allowable range has not yet intersected with that layer, it indicates that the local geometric change in step one can change whether the printing layer produces an actual cured area, and the corresponding position is recorded as a manufacturing influence area related to the cross-section's existence status. For small structures that exist in all three states and remain connected, the narrowest continuous entity channel between the two main bodies on either side of the connection is found within the connection area, and this channel is compared with the stable forming conditions in the current material configuration of the equipment. When some allowable geometric states can meet the stable forming conditions, while other states have crossed the stable forming boundary, this position is recorded as a manufacturing influence area related to fine connection or thin-wall forming. For hollow toys, shell structures, grooves, and locations where accessories may form locally enclosed spaces with the main body, check whether non-solid areas in the slice can be connected to the outside of the model through continuous blank paths. If the current state remains open, and the geometry corresponding to a certain allowable boundary causes the solids on both sides of the opening to close and form an internal area completely surrounded by the solid outline, then this location is recorded as the manufacturing influence area related to the open / closed state. The above judgments all directly use the current surface and local allowable geometry in the 3D reconstruction model from step one, and compare them at the same printing layer and the same local location, so that each manufacturing influence can be traced back to the specific reconstruction uncertainty area.

[0035] In one specific embodiment, for regions where the connection relationships, local entity existence states, fine connection formation states, and open / closed states remain consistent, the changes in the local curing range caused by permissible geometric changes are compared. Using the local entity cross-section formed on the current 3D surface in this printing layer as a reference, the increase or decrease in the entity cross-section caused by the geometric states corresponding to the inner and outer permissible boundaries is observed and compared with the normal cross-section variation range determined for the current material in the equipment process configuration. When both constraint states fall within the normal variation range, the region is recorded as a normal geometric change; when a constraint state causes the entity cross-section to cross the normal variation range, the location is recorded as a manufacturing influence region related to the curing range. This process is particularly suitable for large-area continuous curved surfaces such as the back, abdomen, and head of toys, because the geometric positions of these areas may vary to some extent. However, as long as the connection relationships, internal states, and actual curing range remain within a stable manufacturing range, they will not be confused with the same manufacturing influence as areas such as the base of fingers and ear connections that could truly change the layer-by-layer forming result.

[0036] After completing the layer-by-layer analysis of all local permissible geometric ranges, the resulting manufacturing effects are directly written back to the corresponding spatial region of the 3D reconstructed model. For each manufacturing effect region, its original current 3D surface and local permissible geometric range are retained, and the actual changed manufacturing effect type and the position of the continuous layers covered by this effect in the printing direction are associated. If the entity connection relationship changes first in the same local region, accompanied by a change in the local curing range, then the connection state change is taken as the main manufacturing effect type for that region; if the entity topology remains consistent while the fine connection crosses the stable forming condition, it is recorded as a manufacturing effect related to the fine connection forming; only when the aforementioned structural states remain consistent is the curing range change type used. The manufacturing effect types formed in this way have a clear manufacturing sequence relationship, allowing the next step to directly organize printing events based on the effect type and the position of the continuous layers without re-analyzing the 3D geometry.

[0037] The final manufacturing constraint model is formed. This model fully inherits the current 3D surface, the determined 3D surface, and the local permissible geometric range corresponding to each under-observed region from the S1 3D reconstruction model. It also adds the manufacturing influence region, manufacturing influence type, and corresponding continuous printing layer position at the corresponding locations where manufacturing states are crossed. Thus, the question obtained in step one, "how the local surface is allowed to change under the existing acquisition conditions," is further transformed into "whether this permissible change will alter the actual layer-by-layer manufacturing state," while maintaining a direct correspondence with the original 3D geometry.

[0038] In this embodiment, specifically in step S3, the current three-dimensional surface, the determined three-dimensional surface, the local permissible geometric range, the manufacturing influence area, the manufacturing influence type, and the corresponding continuous printing layer position are first read. The actual printing task is then generated according to the same printing direction and base layer thickness used in step two. The current three-dimensional surface is used to form the layer-by-layer solid cross-section that ultimately enters the printing device; the determined three-dimensional surface is used to maintain the spatial correspondence between the manufacturing event and the specific object structure; the local permissible geometric range is used to determine the range within which the manufacturing influence may appear earlier or end later in the printing direction; the manufacturing influence area is used to locate the affected local entities within each printing layer; the manufacturing influence type is used to determine the printing parameter category corresponding to the event; and the continuous printing layer position is directly used as the main layer segment in the manufacturing event where a manufacturing state change has been confirmed. Thus, the manufacturing constraint information formed in step two is reorganized in this step according to the actual printing sequence, enabling the manufacturing influence in three-dimensional space to be transformed into manufacturing events that the printing device can read layer by layer.

[0039] In one specific embodiment, a complete printable slice is first generated using the current 3D surface in the manufacturing constraint model. Slice planes are arranged sequentially according to a predetermined printing direction, ensuring the spacing between adjacent slice planes matches the base layer thickness of the target 3D printing device's current task. The intersection line between each slice plane and the current 3D surface is calculated, and the area inside the closed intersection line is identified as the actual solid region to be formed in that printable layer. Internal holes, multiple independent solid regions, and the connection relationships between local solids and the main structure are preserved. Since the manufacturing impact judgment in step two uses the same printing direction and base layer thickness, the positions of the continuous printable layers recorded in step two can directly correspond to the actual slices formed here. Subsequently, the intersection of each manufacturing impact region and the corresponding printable layer's slice plane is calculated. The local solid range where the intersection occurs is taken as the location of the manufacturing event for that layer, and the 3D surface is used to determine which stable structure of the object this location specifically belongs to. For example, in step two, it is determined that the connection state of the toy's right arm root changes in several consecutive printing layers. In this step, the local cross-sections where the right arm connects to the torso are extracted in the same layer position, and these local cross-sections are continuously anchored to the defined structural combination of the right arm and torso. If there are also fine connection forming changes at the left ear root in the same printing layer, the left ear event maintains a spatial correspondence separately according to the defined structure of the left ear and the head. This allows multiple local manufacturing effects appearing in the same layer to enter the printing task according to the actual object structure.

[0040] In this process, the main layer segment of the manufacturing event is formed at the continuous printing layer position determined by S2. This main layer segment represents the continuous printing range where manufacturing state changes have been confirmed between the current 3D surface, the geometric state corresponding to the inner allowable boundary, and the geometric state corresponding to the outer allowable boundary. For the start and end of the main layer segment, the positional changes of the corresponding local allowable geometric range along the printing direction in the manufacturing constraint model are further used to determine the entry and exit segments of the manufacturing event. The calculation adopts the conversion relationship between the continuous spatial change range and the fixed slicing interval during discrete slicing, that is, the maximum positional change of the local allowable geometric range relative to the current 3D surface in the current boundary direction is divided by the base layer thickness and rounded up, so that the allowable change range can be covered by the complete printing layer. ; in, This indicates the number of print layers that need to be included in the manufacturing event beyond the current boundary of the main layer segment; This represents the maximum positional change of the corresponding manufacturing influence area relative to the current 3D surface, formed by the local allowable geometry, in the current analysis direction. This data is directly read from the manufacturing constraint model. This indicates the base layer thickness used in the current printing task, which is read from the current task configuration of the photopolymer 3D printing device. and The ratio of these two values, representing similar spatial scales, corresponds to the number of print layers that the allowed geometric changes may span. Rounding up yields the number of print layers that can completely cover the change. For the beginning of a main layer segment, the entry segment is determined by the maximum positional change extending forward from the local allowed geometric range in the print sequence; for the end of a main layer segment, the exit segment is determined by the maximum positional change extending backward from the local allowed geometric range in the print sequence. When there is no allowed positional change beyond the main layer segment in a certain direction, no event coverage layer is added in that direction. After forming the entry and exit segments, each newly added print layer is checked to see if the actual slices still contain the local structure corresponding to the manufacturing influence area. When a newly added print layer no longer has a corresponding slice entity, the manufacturing event in that direction terminates at the previous print layer where the local structure still exists, ensuring that the event coverage is limited by both the local allowed geometric range and the actual model structure's range.

[0041] In one specific embodiment, for example, step two has confirmed that the connection state change at the root of a toy arm continuously covers seven printed layers in the main body segment. The manufacturing constraint model indicates that the maximum allowable positional change of this connection area in the direction of entry of the main body segment is equivalent to 1.6 base layer thickness. and The ratio is 1.6, rounded up to obtain two printed layers. Therefore, two printed layers are first added before the main layer segment as an entry segment. If the actual slice of the first newly added printed layer no longer contains the local entity corresponding to the manufacturing influence area of ​​the arm root, the entry segment ends at the next printed layer where the structure still exists. If the maximum allowable positional change in the exit direction of the main layer segment is equivalent to one base layer thickness, a printed layer is added after the main layer segment as an exit segment, and the same method is used to confirm that the corresponding local structure still exists in this printed layer. The resulting entry segment, main segment, and exit segment together constitute the complete printed layer range of the arm root connection manufacturing event.

[0042] After the event boundaries are determined, the local influence areas of the same manufacturing influence region in consecutive printing layers are organized into complete manufacturing events according to spatial continuity. For two adjacent layers, it is first determined whether the local influence areas in the two layers correspond to the same defined three-dimensional surface structure, and then it is determined whether the local influence area of ​​the upper layer extends along the printing direction and continuously overlaps with the local influence area of ​​the lower layer; if the continuity is satisfied, the two layers are classified into the same manufacturing event. The manufacturing influence type retains the type determined in step two, and each layer retains its own local influence area according to the printing layer sequence. For example, if the base of a toy ear gradually transitions from a narrower connection to a wider connection, as long as step two still determines the relevant consecutive layers as a fine connection forming change, and each layer belongs to the same ear and head connection structure, then a continuous ear base manufacturing event is formed; each printing layer within the event still retains its own local influence area in the actual slice, so that subsequent printing can correspond to the actual cross-sectional changes. For cases where the same local structure exhibits changes in manufacturing influence type in different layers, adjacent events are established sequentially according to the manufacturing influence results formed in step two. For example, if the front layer of the arm root exhibits a fine connection forming change, and the subsequent layer further exhibits a connection state change, then the previous layer forms a fine connection manufacturing event, and the next layer forms a connection state manufacturing event. The two events maintain the spatial relationship of the same arm root structure and are arranged continuously according to the actual printing order.

[0043] In one specific embodiment, for each established manufacturing event, the corresponding printing parameter category is written according to the manufacturing impact type in step two. Changes in connection state correspond to connection structure adaptation parameter categories, used in the next step to select printing parameters that are conducive to maintaining a stable connection between the main body and the local structure; changes in fine connection forming correspond to fine structure adaptation parameter categories; the appearance or disappearance of a cross-section corresponds to local newly formed structure adaptation parameter categories; changes in open / closed state correspond to closed structure adaptation parameter categories; and changes in curing range correspond to cross-sectional change adaptation parameter categories. The parameter category serves as a process call index in the printing task, establishing a correspondence with the parameter combinations already configured for the current material by the current printing device. Taking a connection state event as an example, the printing task records that the event belongs to the connection structure adaptation category, and the next step can read the corresponding exposure and separation parameters in the current material configuration of the device based on this category; taking an open / closed state event as an example, the corresponding closed structure adaptation category is recorded, enabling subsequent execution processes to call parameter combinations suitable for the layer-by-layer formation of this structure.

[0044] Subsequently, an execution record is created for each printing layer according to the actual printing sequence. Each execution record stores the actual slice data generated from the current 3D surface; printing layers outside the manufacturing event record the normal printing state; printing layers within the manufacturing event range simultaneously record the manufacturing influence type, the local position of the layer, the defined main structure to which it belongs, the corresponding parameter category, and the stage of the layer in the current event. Based on the aforementioned event boundary calculation results, the manufacturing event is divided into an entry segment, a main segment, and an exit segment: the entry segment covers printing layers where the manufacturing influence may appear earlier based on the local allowable geometry; the main segment corresponds to the continuous printing layers where the manufacturing state has been confirmed to have been crossed in step two; and the exit segment covers subsequent printing layers where the local allowable geometry and the continuity relationship between adjacent layers indicate that the manufacturing influence may still continue. For printing layers that simultaneously contain multiple independent local events, each event stores its own local range and type, thereby enabling the printing task to express the actual situation that "different spatial locations of the same layer are subject to different manufacturing influences."

[0045] In a specific embodiment, taking a customized toy quickly reconstructed by a user from multi-view photographs as an example, step two has already determined the connection state manufacturing influence at the root of the right arm and the fine connection forming manufacturing influence at the root of the left ear, while retaining the local allowable geometry of a poorly observed area on the back where no manufacturing influence has formed. After generating actual slices based on the current 3D surface, the back area enters the printing task according to the normal printing layer; the root of the right arm forms the main event segment based on the continuous printing layer positions recorded in step two, and then determines the entry and exit segments based on the maximum positional change that the local allowable geometry of this area can reach along the printing direction, and uses actual slices to confirm that the local structure corresponding to the root of the right arm still exists in the expanded printing layer, thus forming a complete right arm connection manufacturing event. When the printing process continues to the corresponding position of the left ear, the fine connection manufacturing event of the left ear is formed in the same way, so that different uncertain reconstruction areas form printing events corresponding to their own spatial positions and manufacturing influences.

[0046] The final output is the printing task. The printing task contains complete layer-by-layer slice data formed according to the current 3D surface, and associates manufacturing event information in the corresponding printed layers affected by manufacturing. The main layer segment of the manufacturing event is directly derived from the continuous printed layer positions recorded in step two, the local action area is derived from the manufacturing influence area, the event type is derived from the manufacturing influence type, the spatial structure is determined by the defined 3D surface, and the entry and exit segments are determined by the local allowable geometric range combined with the continuity relationship of actual adjacent slices.

[0047] In this embodiment, specifically in step S4, the printing task has already saved complete slice data and manufacturing event information associated with the corresponding printing layer according to the actual printing order. This includes the type of manufacturing impact, the location of local action, the defined main structure to which it belongs, the event entry segment, the main segment, the exit segment, and the parameter category. After the 3D printing equipment loads the printing task, it first reads the conventional printing parameters according to the equipment process configuration corresponding to the current material, and establishes the correspondence between each manufacturing event parameter category and the pre-calibrated adaptation parameter combination of the equipment. Conventional printing parameters are used for ordinary printing layers. Changes in connection state correspond to the connection structure adaptation parameter combination, changes in fine connection forming correspond to the fine structure adaptation parameter combination, the appearance or disappearance of the cross section corresponds to the local newly formed structure adaptation parameter combination, changes in open and closed states correspond to the closed structure adaptation parameter combination, and changes in curing range correspond to the cross section change adaptation parameter combination. The aforementioned combination of adaptation parameters is pre-calibrated and saved by the target 3D printing equipment using standard structural specimens under current material conditions. For example, when printing specimens of different connection sizes for thin-joint structures, exposure and motion parameters that ensure continuous and complete forming and stable separation processes are selected as adaptation parameters for thin structures. When printing specimens of different cross-sections for structures with rapidly expanding cross-sections, parameters that ensure stable interlayer separation are selected as adaptation parameters for cross-section changes. The parameter categories in the printing task are only responsible for calling the process configurations that have already been established with the current material and equipment, enabling manufacturing events to be directly converted into actual execution parameters.

[0048] After printing begins, the equipment performs forming layer by layer according to the slice order in the printing task. For ordinary printing layers not associated with manufacturing events, the solid cross-section of the layer is read and exposure, curing, separation, and interlayer movement are completed using conventional printing parameters. For printing layers entering the scope of manufacturing events, the manufacturing impact type, local action position, and main structure of the manufacturing event in that layer are read first, and the corresponding adaptation parameter combination is called according to the parameter category. The event entry segment is used to switch the equipment to the corresponding forming state before the arrival of the manufacturing-sensitive main layer segment determined in step two. For example, if the connection state event of the right arm root already has an entry segment before the main segment, the connection structure adaptation parameters are called from the first layer of the entry segment, and the corresponding forming is performed according to the local action position of the right arm root recorded in that layer. After entering the main segment, printing continues according to the parameter category of the same manufacturing event, while the current slice and the local action position of the current layer are reread for each layer, so that the adaptation action changes layer by layer with the real three-dimensional structure. For example, the cross-section of the arm root may gradually increase with the printing height, and the equipment still uses the actual connection area recorded in the current layer as the action position, rather than using the fixed area of ​​the first layer of the event. Upon reaching the event exit segment, the corresponding adaptation parameters are maintained until the last layer of the exit segment is completed, at which point the normal printing parameters are restored from the next normal layer. In this way, the event entry and exit boundaries determined by the local allowable geometry in step three are directly converted into the device parameter switching moments.

[0049] In one specific embodiment, for connection state change events, the printing control focuses on maintaining the continuous forming between the local structure and the defined main structure. Taking a photopolymerization printing device as an example, the parameter combination corresponding to the connection event can use the device's pre-calibrated connection structure exposure conditions and relatively gentle separation motion conditions to gradually establish a stable connection at the root of the arm, the root of the ear, or the attachment connection in several consecutive layers. When the device supports local exposure control, based on the local action position saved in the printing task, the connection structure exposure configuration is only called in the area corresponding to the manufacturing event, while the remaining solid areas continue to use the current layer's conventional exposure configuration; when the device uses a unified exposure method for the entire layer, the entire layer connection structure adaptation configuration is called in the printing layer where the event occurs. For fine connection forming events, the processing object is the fine connection, thin wall, or narrow connection channel identified in step two. The device calls the fine structure parameter combination corresponding to the current material to maintain the continuous forming of the structure in the entry section, main body section, and exit section. For example, the base of a toy's ear gradually expands from narrow connections to stable connections across several layers. The printing task has recorded the actual position of the ear base in each layer. The device sequentially completes the corresponding slices and continuously calls the fine structure adaptation parameters until the exit segment is completed.

[0050] In one specific embodiment, for events involving the appearance or disappearance of a cross-section, the device employs local nascent structure adaptation parameters starting from the event entry segment. This allows small cross-section structures such as fingertips, ear tips, and decorative protrusions to adopt a suitable process state for their formation at the earliest possible entry point into the printing layer, and maintain this state within the main body segment. As the local structure gradually merges into a larger entity or completely exits the slice, the device reverts to normal printing parameters at the end of the exit segment. For events involving changes in open / closed states, the device, based on the type of closed structure parameters in the printing task, calls the corresponding parameter combination in continuous printing layers where internal channels gradually close or local cavities gradually form. This ensures that the opening shrinks, the closure forms, and subsequent entity coverage is completed continuously according to the same manufacturing event. For example, if a hollow toy has a gradually closing local opening on its back, and step three has already determined the entry segment, main body segment, and exit segment of this closing event, the device calls the closed structure adaptation parameters starting from the entry segment until the local closed structure is completed and leaves the manufacturing sensitive layer segment. In the event of a change in the curing range, the equipment calls upon the cross-sectional change adaptation parameters to ensure that the layers in which the curing area of ​​the neck extends significantly towards the head, or the slender limbs extend towards the larger body, adopt the exposure and separation conditions that the equipment has calibrated for large cross-sectional changes, so as to maintain the stable formation of the layer and its adjacent layers.

[0051] When multiple manufacturing events exist within the same printing layer, the entity region corresponding to each event is determined based on the local impact location recorded in the printing task. When the photopolymer 3D printing equipment supports local exposure control, each local region calls the exposure configuration corresponding to its own manufacturing influence type, while the remaining regions execute according to the layer's regular exposure configuration. For parameters that require unified execution across the entire layer, such as lifting speed and separation motion, the unified motion configuration for this layer is determined according to a pre-determined priority order of manufacturing states. The priority order is as follows: connection state change, open / closed state change, cross-section appearance or disappearance, fine connection forming change, and curing range change. When the former type exists, its corresponding whole-layer motion parameters are used, while subsequent types retain their corresponding local exposure configurations. This order is set according to the degree of direct influence of different manufacturing states on the entity topology and continuous forming process, ensuring that a unique whole-layer motion configuration can be determined when multiple manufacturing events exist simultaneously within the same printing layer. For example, if there is a connection state change at the root of the right arm and a fine connection forming change at the root of the left ear in the same printing layer, the entire layer adopts the separation motion configuration corresponding to the connection state change, while the root of the right arm and the root of the left ear call their respective local exposure configurations according to their manufacturing events.

[0052] In one specific embodiment, for example, for a customized toy rapidly reconstructed from multi-view photographs, in the printing task output in step three, the torso is a normal printing layer, followed by a connection event at the base of the right arm and a fine connection forming event at the base of the left ear above the head. The device first completes the torso according to the normal printing parameters; when it reaches the entry segment of the right arm connection event, it switches to the connection structure adaptation parameters, continuously completing the entry segment, main body segment, and exit segment according to the actual position of the right arm base recorded in each layer, and then resumes the normal printing state; when printing continues to the left ear event, it switches to the fine structure adaptation parameters, and continuously forms the cross-section that changes layer by layer with the base of the ear, and after exiting this event, it resumes the normal parameters again. The back area, which has been determined to be a normal geometric change in step two, continues to be formed directly according to the normal slices in the printing task. The 3D printing device executes all printing layers in this order until all slices corresponding to the current 3D surface are manufactured, and then completes the solid processing according to the predetermined separation, cleaning, and post-curing processes of the current material and equipment, finally obtaining the 3D printed entity. The 3D printed entity is the output of this step, completing the entire technology chain from rapid 3D reconstruction, manufacturing impact recognition, manufacturing event generation to actual adaptive printing.

[0053] like Figure 2 As shown in the figure, this invention also provides a rapid reconstruction and adaptive printing system for three-dimensional models, the system comprising: The 3D reconstruction processing unit 10 is used to acquire object acquisition data of the target object to establish the current 3D surface, including determining the current surface of the 3D surface and the insufficient observation area. For the current surface of the insufficient observation area, the local allowable geometric range is determined based on the current surface to generate a 3D reconstruction model. The 3D reconstruction model includes determining the 3D surface, the current surface of the insufficient observation area, and the corresponding local allowable geometric range. The current surface of the insufficient observation area is determined based on the area where there are local geometric changes. The manufacturing impact analysis unit 20 is used to determine, based on the three-dimensional reconstruction model, three corresponding local solid sections of the current three-dimensional surface and the local allowable geometry within the same printing layer along the printing direction, wherein the three-dimensional surface portion is determined to be consistent; and to compare the layer-by-layer manufacturing state among the three local solid sections within the same local allowable geometry. When any geometric state undergoes a manufacturing state change on any affected printing layer, the location corresponding to the local allowable geometric range is marked as the manufacturing influence area, and the manufacturing influence type and the location of the continuous printing layer are recorded. At the same time, it is associated with the three-dimensional reconstruction model to form a manufacturing constraint model; the geometric state corresponding to the inner allowable boundary and the geometric state corresponding to the outer allowable boundary within the local allowable geometric range. The printing task generation unit 30 is used to generate layer-by-layer slicing data according to the manufacturing constraint model, with the same printing direction and base layer thickness as S2, map the manufacturing influence area to the corresponding printing layer, take the continuous printing layer position as the main layer segment, combine the local allowable geometric range to determine the entry segment and exit segment, form a manufacturing event and associate it with the printing parameter category to obtain manufacturing event information, so as to construct a printing task; the printing task includes layer-by-layer slicing data and manufacturing event information; The printing control unit 40 is used to read the printing task, determine the corresponding equipment adaptation parameter combination according to the manufacturing event information, call the conventional parameters in the normal printing layer, and call the equipment adaptation parameter combination in the entry segment, main segment and exit segment of the corresponding manufacturing event to print layer by layer to obtain the printing results of each layer; after completing all printing layers, the results are summarized to obtain the three-dimensional printed entity.

[0054] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0055] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of apparatuses or units may be electrical, mechanical, or other forms.

[0056] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0057] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A method for rapid reconstruction and adaptive printing of 3D models, characterized in that, The method includes: S1. Acquire object acquisition data of the target object to establish the current three-dimensional surface, including determining the current surface of the three-dimensional surface and the insufficient observation area. For the current surface of the insufficient observation area, determine the local allowable geometric range based on the current surface to generate a three-dimensional reconstruction model. The three-dimensional reconstruction model includes determining the current surface of the three-dimensional surface and the insufficient observation area, as well as the corresponding local allowable geometric range. The current surface of the insufficient observation area is determined based on the region where there are local geometric changes. S2. Based on the aforementioned 3D reconstruction model, determine three corresponding local solid sections along the printing direction that correspond to the current 3D surface and the local allowable geometry within the same printing layer, wherein the 3D surface portion is kept consistent; compare the layer-by-layer manufacturing state among the three local solid sections within the same local allowable geometry. When any geometric state undergoes a manufacturing state change on any affected printing layer, the location corresponding to the local allowable geometric range is marked as the manufacturing influence area, and the manufacturing influence type and the location of the continuous printing layer are recorded. At the same time, it is associated with the three-dimensional reconstruction model to form a manufacturing constraint model; the geometric state corresponding to the inner allowable boundary and the geometric state corresponding to the outer allowable boundary within the local allowable geometric range. S3. Based on the manufacturing constraint model, generate layer-by-layer slicing data according to the same printing direction and base layer thickness as in S2, map the manufacturing influence area to the corresponding printing layer, take the continuous printing layer position as the main layer segment, combine the local allowable geometric range to determine the entry segment and exit segment, form manufacturing events and associate them with printing parameter categories to obtain manufacturing event information, so as to construct a printing task; the printing task includes layer-by-layer slicing data and manufacturing event information. S4. Read the printing task and determine the corresponding equipment adaptation parameter combination based on the manufacturing event information; call the regular parameters in the ordinary printing layer, and call the equipment adaptation parameter combination in the entry segment, main segment and exit segment of the corresponding manufacturing event to print layer by layer to obtain the printing results of each layer; after completing all printing layers, summarize to obtain the three-dimensional printed entity.

2. The method for rapid reconstruction and adaptive printing of a 3D model according to claim 1, characterized in that, The acquisition of object collection data for the target object includes: When using multi-view photographs, keep the target object still, acquire images from multiple viewing directions, and ensure that there are repeatable observation areas of the object surface between adjacent viewpoints; A correspondence is established based on the common visible object features in adjacent images. The corresponding features are used to obtain the position and direction of each shooting perspective relative to the coordinate relationship of the same object, so that the spatial positions obtained from different perspectives are unified into the same three-dimensional reconstructed coordinate relationship.

3. The method for rapid reconstruction and adaptive printing of a 3D model according to claim 1, characterized in that, The determination of the local permissible geometry based on the current surface includes: Extract the established 3D surface boundaries around the currently under-observed region, and generate a continuous surface towards the missing region along the extension direction of the original surface at the boundary. The position of the surface to be verified is moved step by step along the inner and outer directions of the current local surface, and the object acquisition data is used to determine whether the position still matches the actual observation. When two adjacent attempts correspond to satisfying and exceeding the existing observation constraints, respectively, the search range is further narrowed between the two until the local boundary position is stable, thus forming the inner and outer allowable boundaries of the current local region.

4. The method for rapid reconstruction and adaptive printing of a 3D model according to claim 1, characterized in that, The comparison of layer-by-layer manufacturing states includes determining the entity connectivity state: The interior of the closed contour obtained by slicing is taken as the solid region, and the same connectivity label is assigned to regions that are in contact with each other and can be connected by continuous solid paths. When a local structure in the current section is connected to the main body, but the two are separated in the section corresponding to the inner allowable boundary or the section corresponding to the outer allowable boundary, the local position is recorded as the manufacturing influence area where the connection state changes. When a local structure in the current section separates from the main body, and a new solid connection is formed between the section corresponding to the inner allowable boundary or the section corresponding to the outer allowable boundary, it is also recorded as a manufacturing influence area where the connection state changes.

5. The method for rapid reconstruction and adaptive printing of a three-dimensional model according to claim 1, characterized in that, The comparison of layer-by-layer manufacturing states includes determining whether a local entity appears or disappears: For local structures extending outward from the main body, check layer by layer whether the corresponding entity regions in the current section, the section corresponding to the inner allowable boundary, and the section corresponding to the outer allowable boundary exist simultaneously; When a local solid has been formed on a certain 3D surface in a certain printing layer, and the geometric state corresponding to the inner allowable boundary or the outer allowable boundary has not yet intersected with the printing layer, it is recorded as a manufacturing influence area related to the cross-section existence state.

6. The method for rapid reconstruction and adaptive printing of a three-dimensional model according to claim 1, characterized in that, The determination of the entry and exit segments based on the locally permissible geometric range includes: The number of additional printing layers required is determined based on the maximum positional variation of the local permissible geometric range along the printing direction; Add corresponding entry and exit segments before and after the main layer segment, and check layer by layer whether the local structure corresponding to the manufacturing impact area still exists in the newly added printing layer; When a newly added printing layer does not have a corresponding local structure, the manufacturing event in the current direction terminates at the previous printing layer where the local structure still exists.

7. The method for rapid reconstruction and adaptive printing of a three-dimensional model according to claim 6, characterized in that, The events that form the manufacturing event include: The local locations of the same manufacturing influence area within consecutive printing layers are organized into manufacturing events according to their spatial continuity. When the local action positions in two adjacent layers correspond to the same defined three-dimensional surface structure, and the local action position of the upper layer extends along the printing direction and continuously overlaps with the lower layer, the two adjacent layers are classified into the same manufacturing event. The manufacturing impact type remains the established type, and the local scope of each layer is preserved according to the printing layer order.

8. The method for rapid reconstruction and adaptive printing of a three-dimensional model according to claim 1, characterized in that, The associated printing parameter categories include: The printing parameter categories are determined based on the type of manufacturing impact; among them, changes in connection state correspond to the connection structure adaptation parameter category, changes in fine connection forming correspond to the fine structure adaptation parameter category, the appearance or disappearance of cross-sections correspond to the local new structure adaptation parameter category, changes in open and closed states correspond to the closed structure adaptation parameter category, and changes in curing range correspond to the cross-section change adaptation parameter category.

9. The method for rapid reconstruction and adaptive printing of a three-dimensional model according to claim 1, characterized in that, The step of calling the device adaptation parameter combination for layer-by-layer printing in the entry, main, and exit segments of the corresponding manufacturing event includes: When multiple manufacturing events exist in the same printing layer, the entity region corresponding to each event is determined according to the local action position recorded in the printing task record; For printing devices that support local exposure control, each local area calls the exposure configuration corresponding to its own manufacturing effect type, while the remaining areas execute according to the normal exposure configuration of that layer; For parameters that need to be executed uniformly across the entire layer, the uniform motion configuration of this layer is determined according to the predetermined priority order of manufacturing states. The priority order is as follows: connection state change, open and closed state change, cross-section appearance or disappearance, fine connection forming change, and curing range change. Specifically, the device adaptation parameter combination is invoked at the first layer of the entry segment of the manufacturing event and continues until the main segment and the exit segment; after the last layer of the exit segment is completed, the normal printing parameters are restored from the next normal printing layer.

10. A rapid reconstruction and adaptive printing system for three-dimensional models, characterized in that, The system includes: A 3D reconstruction processing unit is used to acquire object acquisition data of the target object to establish the current 3D surface, including determining the current surface of the 3D surface and the current surface of the under-observed region. For the current surface of the under-observed region, the local allowable geometric range is determined based on the current surface to generate a 3D reconstruction model. The 3D reconstruction model includes determining the 3D surface, the current surface of the under-observed region, and the corresponding local allowable geometric range. The current surface of the under-observed region is determined based on the region where there are local geometric changes. The manufacturing impact analysis unit is used to determine, based on the 3D reconstruction model, three corresponding local solid sections of the current 3D surface and the local allowable geometry within the same printing layer along the printing direction, wherein the 3D surface portion is determined to be consistent; and to compare the layer-by-layer manufacturing state among the three local solid sections within the same local allowable geometry. When any geometric state undergoes a manufacturing state change on any affected printing layer, the location corresponding to the local allowable geometric range is marked as the manufacturing influence area, and the manufacturing influence type and the location of the continuous printing layer are recorded. At the same time, it is associated with the three-dimensional reconstruction model to form a manufacturing constraint model; the geometric state corresponding to the inner allowable boundary and the geometric state corresponding to the outer allowable boundary within the local allowable geometric range. The printing task generation unit is used to generate layer-by-layer slicing data according to the manufacturing constraint model, with the same printing direction and base layer thickness as S2, map the manufacturing influence area to the corresponding printing layer, take the continuous printing layer position as the main layer segment, combine the local allowable geometric range to determine the entry segment and exit segment, form a manufacturing event and associate it with the printing parameter category to obtain manufacturing event information, so as to construct a printing task; the printing task includes layer-by-layer slicing data and manufacturing event information. The printing control unit is used to read the printing task, determine the corresponding equipment adaptation parameter combination based on the manufacturing event information, call the conventional parameters in the normal printing layer, and call the equipment adaptation parameter combination in the entry segment, main segment and exit segment of the corresponding manufacturing event to print layer by layer to obtain the printing results of each layer; after completing all printing layers, the results are summarized to obtain the three-dimensional printed entity.

Images