A depth-optimized three-dimensional color control method based on color printing materials

Abstract

This invention provides a depth-optimized 3D color control method based on color printing materials, comprising: constructing a primary color material system, classifying and assigning it to corresponding primary color channels according to the transparency properties of the printing materials; performing voxelization on the 3D model to be printed, and dividing the object surface voxels into multiple logical layers corresponding to the number of primary color channels inward along the normal direction based on a directed distance field; independently performing 3D halftone screening within each logical layer to generate an initial voxel binary distribution for each primary color channel; based on a physical optics model, using depth units formed along the surface voxel normal direction as the basic evaluation unit, keeping the total volume coverage of each primary color channel material constant, iteratively optimizing the initial voxel binary distribution through local voxel exchange operations, and using the resulting voxel distribution for 3D color printing. This invention can control the optical transmission path of multi-layer semi-transparent materials, significantly improving the accuracy of 3D printing color reproduction.

Description

Technical Field

[0001] This invention relates to the fields of additive manufacturing data processing and digital image processing technology, and in particular to a depth-optimized three-dimensional color control method based on color printing materials. Background Technology

[0002] Multi-material inkjet 3D printing technology enables high spatial resolution and color expression freedom by precisely controlling the deposition positions of multiple materials at the voxel scale. Existing technologies typically convert target color information into multi-material printing instructions through voxel modeling and slice-by-slice data processing, achieving the shaping and rendering of colored 3D objects to a certain extent. However, in high-fidelity color 3D printing scenarios, existing methods still face several technical bottlenecks.

[0003] Current mainstream inkjet 3D printing typically uses colored and colorless photosensitive polymers with varying transparency as output materials. The colored photosensitive polymers used usually possess a degree of translucency, causing incident light to undergo not only absorption and superposition within the material but also scattering and multiple propagation—volume optical behaviors. Under these conditions, traditional methods based on surface coloring or single-layer color mapping struggle to accurately describe the comprehensive impact of material stacking on color reproduction, easily leading to issues such as reduced color saturation, narrowed effective color gamut, and blurred surface details in the printed product.

[0004] To meet the computational efficiency requirements of large-scale voxel data, existing technologies generally employ localized or hierarchical data processing and optimization strategies. However, in practical applications, these strategies often fail to explicitly introduce collaborative modeling and targeted control of the optical effects of material stacking, making it difficult to fully explore and utilize the optical properties of multi-material systems while ensuring computational efficiency. Therefore, how to introduce an optical perception optimization mechanism oriented towards material stacking structures based on the advantages of high-efficiency computing has become a key technical issue for improving the color reproduction accuracy and visual consistency of color 3D printing. Summary of the Invention

[0005] Therefore, it is necessary to provide a depth-optimized three-dimensional color control method based on color printing materials to address the aforementioned technical problems.

[0006] A depth-optimized 3D color control method based on color printing materials includes the following steps:

[0007] S1, Construct a primary color material system, wherein the primary color material system is classified according to the transparency properties of the printing material and assigned to the corresponding primary color channels;

[0008] S2, the 3D model to be printed is voxelized, and the voxels on the surface of the object are divided into multiple logical layers corresponding to the number of primary color channels inward along the normal direction based on the directed distance field. Each logical layer corresponds to one primary color channel.

[0009] S3, within each logic layer, independently performs three-dimensional halftone screening to generate the initial voxel binary distribution of each primary color channel;

[0010] S4, based on the physical optics model, uses the depth unit formed along the surface voxel normal direction as the basic evaluation unit. Under the premise of keeping the total volume coverage of each primary color channel material unchanged, the initial voxel binary distribution is iteratively optimized through local voxel exchange operation until the preset conditions are met, and the optimized voxel distribution is obtained for three-dimensional color printing.

[0011] In one embodiment, dividing the object surface voxels into multiple logical layers corresponding to the number of primary color channels inward along the normal direction based on a directed distance field includes:

[0012] Calculate the directed distance from each voxel in the voxel mesh to the object surface;

[0013] The layer thickness parameter is determined based on the physical resolution of the printing device, and the layer thickness parameter is equal to the voxel physical size of the printing system in the direction of minimum resolution.

[0014] Based on the layer thickness parameter, a set of equidistant distance thresholds are constructed, and a series of equidistant shells are extracted in the directed distance field using the distance thresholds, dividing the object surface voxels into geometric layers arranged in order corresponding to the number of primary color channels.

[0015] The geometric layers are sequentially mapped to functional layers, and each voxel is assigned a layer index, establishing a mapping relationship from three-dimensional geometric space to multi-layer shading logic.

[0016] In one embodiment, the primary color material system includes opaque materials, translucent materials, and transparent materials, and the functional layer includes a color modulation layer and a reflective substrate layer;

[0017] Among them, the first to N-2 layers from the outside to the inside are color modulation layers, corresponding to semi-transparent or transparent base color materials; the N-1 layer is a reflective base layer, corresponding to non-transparent base color materials.

[0018] In one embodiment, step S3 includes:

[0019] Generate a geometrically adaptive voxel traversal sequence, which is constructed based on the local geometric properties and spatial topological relationships of the model surface;

[0020] Along the voxel traversal sequence, three-dimensional error diffusion processing is performed independently in each logic layer to generate the initial voxel binary distribution of each primary color channel.

[0021] In one embodiment, generating a geometry-adaptive voxel traversal sequence includes:

[0022] Divide the object to be printed into multiple slice layers along the construction direction;

[0023] Identify the connected components within the current slice layer;

[0024] For each surface connected component in the current slice layer, a traversal path is generated using a serpentine traversal strategy or a directed traversal strategy, depending on whether the connected component is newly appearing or about to disappear. The directed traversal strategy includes: determining the traversal starting point and priority traversal direction based on the component of the voxel surface normal in the slice normal direction, and selecting the next voxel during the traversal process based on directional consistency and the distance relationship between the voxel and the empty region.

[0025] In one embodiment, performing the three-dimensional error diffusion process independently within each logic layer includes:

[0026] For the current voxel in the traversal sequence, a local tangent plane and its two-dimensional coordinate system are dynamically constructed based on its surface normal and local traversal direction;

[0027] Map the predefined two-dimensional error diffusion filter to the local tangent plane;

[0028] Within the 3D neighborhood of the current voxel, select a neighboring voxel that matches the filter weight point at the tangent plane projection position and is located after the current voxel in the traversal sequence. Distribute the error generated by the quantization of the current voxel to the neighboring voxel according to the filter weight ratio.

[0029] In one embodiment, based on a depth optimization method, the basic evaluation unit is a depth cell formed along the normal direction of the surface voxels, including:

[0030] For any voxel location on the visible surface of an object, a combination of voxels extending along the surface normal of that voxel and covering the multiple logic layers is defined as a depth unit. The depth unit is used to fully describe the material stacking structure of the surface point along the normal direction.

[0031] In one embodiment, iterative optimization of the initial binary voxel distribution through a local voxel exchange operation includes:

[0032] For each depth cell, the predicted color value is calculated based on the depth optimization method;

[0033] Calculate the color difference between the predicted color value and the target color value, and use it as the color error of the surface voxel corresponding to the depth unit;

[0034] With the goal of minimizing the sum of color errors of all surface voxels, the voxel distribution is iteratively optimized through local voxel exchange operations.

[0035] In one embodiment, calculating the predicted color value based on the depth optimization method includes:

[0036] Optical calculations were performed on the multi-layer material stacked structure within the depth unit using a physical optics model to obtain the comprehensive spectral reflectance.

[0037] The comprehensive spectral reflectance is converted into the International Commission on Illumination (ICI) standard color space value under standard light source conditions and mapped to the ICI uniform color space to obtain the predicted color value;

[0038] The physical optics calculations are accelerated by a pre-built multi-layer color lookup table.

[0039] In one embodiment, the local voxel exchange operation includes:

[0040] Within the same primary color channel layer, with the current voxel as the center, search for candidate voxels located in the same slice and whose binary state is opposite to that of the current voxel within its neighborhood;

[0041] Swap the binary states of the current voxel and the candidate voxel on the primary color channel layer;

[0042] The swap operation is only accepted when the local color difference is reduced after the swap, and the total number of voxels in the primary color channel layer with a state of 1 remains unchanged before and after the swap.

[0043] Compared with existing technologies, the advantages and beneficial effects of this invention are as follows: This invention achieves a structured expression of the overprinted structure of translucent materials by constructing a primary color material system and dividing the logical layers based on a directed distance field; by independently performing three-dimensional halftone screening within each logical layer, it achieves good fit between texture details and complex curved surface shapes, effectively suppressing structural artifacts; by using depth units as the basic evaluation unit and combining physical optics models for local voxel exchange optimization, it achieves effective control of the optical transmission path of multi-layer translucent materials along the normal direction without destroying the total volume coverage of each channel material, significantly improving the accuracy and stability of 3D printing color reproduction and expanding the achievable color performance range. Attached Figure Description

[0044] Figure 1 This is a flowchart illustrating a depth-optimized three-dimensional color control method based on color printing materials in one embodiment.

[0045] Figure 2 This is a schematic diagram of the 3D model output slice layering in one embodiment;

[0046] Figure 3 This is a schematic diagram of the surface system multilayer material coloring in one embodiment;

[0047] Figure 4 This is a schematic diagram of a depth element along the normal direction in one embodiment. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0049] It should be noted that, unless otherwise defined, the technical or scientific terms used in one or more embodiments of this specification should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in one or more embodiments of this specification do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word covers the element or object listed following the word and its equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0050] For ease of understanding, the terms used in the embodiments of this invention are explained below:

[0051] Voxel: A basic discrete unit in three-dimensional space, analogous to a pixel in a two-dimensional image, and the smallest unit for digitizing the three-dimensional model and controlling material deposition in this invention. In this invention, the voxel resolution maintains a 1:1 mapping with the physical resolution of the printing device, and the voxel size is consistent with the physical size in the direction of the minimum resolution of the printing device, used to accurately describe the spatial structure and material distribution of the three-dimensional model.

[0052] Primary color material system: The set of materials used to achieve three-dimensional color printing in this invention is divided into three categories according to the transparency properties of the materials: non-transparent materials, semi-transparent materials and transparent materials. Each type of material corresponds to an independent primary color channel. The optical properties of each channel material are complementary. High-fidelity reproduction of the target color is achieved through superposition and combination. At the same time, material output constraints are set to ensure the rationality of overprinting.

[0053] Primary color channels: Corresponding to each type of material in the primary color material system, these are independent units that implement halftone screening and material allocation control for each channel. In this invention, the number of primary color channels is consistent with the number of logical layers. Each channel corresponds to one logical layer and undertakes a specific color rendering function. Each channel independently performs halftone processing and iterative optimization.

[0054] Directed distance field (SDF): A field used to describe the directed distance from each voxel in the voxel mesh to the surface of the object. The positive or negative value of the distance (SDF value) distinguishes whether the voxel is located inside the model (positive) or outside (negative). The magnitude of the distance value represents the distance of the voxel from the surface. It is the core foundation for realizing the logical layering of surface voxels in this invention and is used to extract equidistant shells and divide geometric layers.

[0055] Logical layering: A layered structure that divides surface voxels inward along the normal direction based on a directed distance field. The number of layers is the same as the number of primary color channels. Each logical layer corresponds to one primary color channel and is divided into a color modulation layer (semi-transparent / transparent material) and a reflective base layer (non-transparent material). It is used to establish the mapping relationship between three-dimensional geometric space and multi-layer shading logic.

[0056] 3D halftone screening: The process of converting the target color value of continuous tone into discrete voxel binary state (deposited / not deposited corresponding primary color material). In this invention, a channel-independent processing method is adopted, combined with geometrically adaptive voxel traversal sequence and 3D error diffusion to ensure color reproduction accuracy and surface detail clarity, and reduce structural artifacts.

[0057] Geometric Adaptive Voxel Traversal Sequence: A voxel traversal path generated based on local geometric properties (such as surface normals) and spatial topological relationships (such as connected component types) of the model surface. It adapts to the surface shape of the model and adopts serpentine or directed traversal strategies for different types of connected components to avoid artifacts caused by conflicts between the traversal direction and the surface shape.

[0058] Error diffusion processing: a core step in halftone screening, used to distribute color errors generated during voxel quantization (binarization) to unprocessed neighboring voxels according to preset filter weights, thereby ensuring uniform error distribution and reducing halftone artifacts. This invention employs a two-dimensional error diffusion filter mapped to a three-dimensional local tangent plane to achieve adaptation to the three-dimensional geometric space, with each logic layer executing independently and errors not propagating across layers.

[0059] Depth cell: A combination of voxels extending along the normal direction of the surface voxels and covering all logic layers. It is the basic evaluation unit for depth iterative optimization in this invention. Each depth cell contains one voxel from each logic layer and fully describes the material stacking structure of the surface points along the normal direction. It is used to calculate color error and perform optimization operations.

[0060] Physical Optical Model: A physical calculation model for describing the light absorption and scattering characteristics of semi-transparent materials. In this invention, it is used to calculate the comprehensive spectral reflectance of multiple layers of materials stacked within a depth unit. The absorption coefficient and scattering coefficient are calculated by combining the reflectance and transmittance of the materials, simulating the multiple scattering and absorption process of light in multi-layer semi-transparent materials, and providing a physical basis for color prediction.

[0061] Multi-layer color lookup table: Based on a pre-built color mapping table of physical optics model, the comprehensive spectral reflectance and CIELAB color value corresponding to different material stacking combinations are calculated in advance. This is used to quickly look up the predicted color value of the depth unit, avoiding the repeated execution of complex optical calculations in each optimization and improving the efficiency of iterative optimization.

[0062] Local voxel exchange operation: The core operation in deep iterative optimization. Within the same primary color channel layer, the current voxel is exchanged with the candidate voxel with the opposite binary state in the neighborhood. The exchange is only accepted under the premise that the local color difference is reduced and the total volume coverage of the material remains unchanged. It is used to fine-tune the voxel distribution and compensate for the color deviation caused by volume optical effects.

[0063] Connected components: Within the same slice layer, a set of voxels that are adjacent to each other and belong to the model to be printed. They are divided into three categories: newly appearing, persistent, and about to disappear. In this invention, different voxel traversal strategies are selected according to the type of connected components to ensure traversal efficiency and path rationality.

[0064] CIE XYZ values: International Commission on Illumination standard color space values. Based on the theory of human eye's trichromatic color perception, they use X, Y, and Z values ​​to represent all visible colors and serve as the basis for conversions to many other color spaces (such as CIELAB).

[0065] CIELAB space: International Commission on Illumination (ICI) uniform color space. A color space based on the characteristics of human color perception, it accurately reflects the differences in color perception. In this invention, it is used to represent target color values ​​and predicted color values. Color difference calculation uses the Euclidean distance of this space as an evaluation standard for color error.

[0066] Volumetric optical effect: The propagation behavior of incident light in semi-transparent printing materials, including light absorption, superposition, scattering and multiple propagation, is the main reason for color reproduction deviation in traditional surface coloring methods. This invention achieves accurate compensation for this effect through physical optical modeling and deep iterative optimization.

[0067] In one embodiment, such as Figure 1 As shown, a depth-optimized 3D color control method based on color printing materials is provided, including the following steps:

[0068] Step S1: Construct a primary color material system. The primary color material system is classified according to the transparency properties of the printing material and assigned to the corresponding primary color channels.

[0069] Specifically, to ensure the stability and controllability of structural forming and color reproduction during 3D color printing, this application first constructs a primary color material system for the inkjet printing system for structural forming and color expression. The primary color material system is classified according to the transparency properties of the printing materials and assigned to corresponding primary color channels. The primary color material system includes three categories: opaque materials, semi-transparent materials, and transparent materials. Materials with different transparency levels correspond to different primary color channels, and each primary color channel undertakes different color presentation functions, laying the foundation for subsequent layering and color control. Among them, semi-transparent and transparent materials are mainly used for color modulation, while opaque materials are mainly used to construct a reflective substrate to ensure light reflection effects and improve the stability of color presentation.

[0070] When constructing a primary color material system, it is necessary to consider the hardware characteristics of the printing equipment, the optical parameters of the materials, and the reproduction requirements of the target colors. This requires rationally dividing the number of primary color channels and their corresponding materials to ensure that the optical properties of each channel's materials are complementary, enabling rich color expression through superposition and combination. Simultaneously, it is necessary to set material output constraints for each primary color channel. For example: transparent and semi-transparent layers can be superimposed to a limited thickness, ideally without affecting the surface shape representation; the same output layer can contain both transparent and semi-transparent materials; once an output layer accepts a non-colored opaque material, transparent or semi-transparent materials are restricted from output on that layer; the chromaticity values ​​of each semi-transparent material should be nearly completely uncorrelated; for semi-transparent layer material output layers, only semi-transparent or transparent materials of the same hue are allowed to be output on the same layer, thereby ensuring the rationality of material overprinting and the stability of optical properties.

[0071] Step S2: Voxelize the 3D model to be printed, and divide the object surface voxels into multiple logical layers corresponding to the number of primary color channels inward along the normal direction based on the directed distance field. Each logical layer corresponds to one primary color channel.

[0072] Specifically, during 3D model output, the continuous 3D model to be printed is first converted into a voxel representation suitable for material distribution using 3D spatial discretization technology. The continuous 3D model is converted into a discrete voxel mesh, with the voxel resolution maintaining a 1:1 mapping to the physical resolution of the printing device, ensuring that the voxel distribution matches the printing accuracy. During the voxelization process, the model shape is divided using an axial parallel bounding box (B), establishing a regular voxel mesh, and classifying voxels into internal voxels, external voxels, and surface voxels. Located in the model The internal set of voxels; external voxels refer to voxels in bounding box B that are not part of the internal set. All voxels, surface voxels The set of voxels that constitute the boundary of an object's surface, that is, those that simultaneously satisfy the condition of being internal voxels. And located on the surface of the object voxels on the surface.

[0073] Based on the Directed Distance Field (SDF), the voxels of the object surface are divided into multiple logical layers along the normal direction inward, corresponding to the number of primary color channels. Each logical layer corresponds to one primary color channel.

[0074] Based on this, the object surface voxels are divided inward along the normal direction into multiple logical layers corresponding to the number of primary color channels, using a directed distance field, including:

[0075] Calculate the directed distance from each voxel in the voxel mesh to the object surface;

[0076] The layer thickness parameter is determined based on the physical resolution of the printing device, and the layer thickness parameter is equal to the voxel physical size of the printing system in the direction of minimum resolution.

[0077] Based on the layer thickness parameter, a set of equidistant distance thresholds are constructed, and a series of equidistant shells are extracted in the directed distance field using the distance thresholds, dividing the object surface voxels into geometric layers arranged in order corresponding to the number of primary color channels.

[0078] The geometric layers are sequentially mapped to functional layers, and each voxel is assigned a layer index, establishing a mapping relationship from three-dimensional geometric space to multi-layer shading logic.

[0079] The primary color material system includes opaque materials, translucent materials, and transparent materials, and the functional layer includes a color modulation layer and a reflective substrate layer;

[0080] Among them, the first to N-2 layers from the outside to the inside are color modulation layers, corresponding to semi-transparent or transparent base color materials; the N-1 layer is a reflective base layer, corresponding to non-transparent base color materials.

[0081] Specifically, the specific division process includes the following sub-steps:

[0082] Calculate the directed distance (SDF value) from each voxel in the voxel mesh to the object surface. Voxels with an SDF value of 0 are defined as geometric surfaces. The sign of the directed distance value can be used to distinguish whether the voxel is located inside or outside the model. The magnitude of the distance value is used to characterize the distance of the voxel from the surface.

[0083] Determine the layer thickness parameters based on the physical resolution of the printing equipment. Layer thickness parameters This is equal to the physical size of the voxels in the direction of minimum resolution of the printing system, ensuring that the layered results are precisely aligned with the voxel grid in spatial topology, avoiding geometric and optical uncertainties caused by thick multi-voxel layers.

[0084] A set of equidistant distance thresholds is constructed based on the layer thickness parameter, i.e. , Using the layer thickness parameter and the distance threshold, a series of equidistant shells are extracted in the directed distance field, dividing the object surface voxels into geometric layers arranged in sequence corresponding to the number of primary color channels. It equals the number of primary color channels N. For the th A layer, whose corresponding voxel set consists of voxels that satisfy the following conditions: voxels The distance value satisfies And at least one voxel exists in its local neighborhood. Make This ensures that each layer is approximately represented as an equidistant shell with a thickness of one voxel in a discrete voxel grid, thereby avoiding the geometric and optical uncertainties caused by multi-voxel thick layers.

[0085] The geometric layers are sequentially mapped to functional layers, and each voxel is assigned a layer index, establishing a mapping relationship from three-dimensional geometric space to multi-layer shading logic. The object's N geometric layers from the outside in are sequentially mapped to N functional layers, named Layer 0, Layer 1, ..., Layer N-1. The arrangement of the layers is as follows: Figure 2 As shown. Among them, Layer 0 to Layer N-2 from the outside to the inside serve as color modulation layers, corresponding to the transparent base color material and the semi-transparent output, respectively, and undertake the main function of color presentation; Layer N-1 serves as the reflective substrate layer, corresponding to the non-transparent base color material, and is used to provide a light reflection substrate to compensate for the light scattering loss of the semi-transparent material.

[0086] In Layers 0 to N-2, voxel states "1" are mapped to the primary colored material corresponding to the primary color channel, while "0" is uniformly mapped to the secondary colored material corresponding to the primary color channel. The primary colored material is responsible for presenting the hue of the primary color in this layer, while the secondary colored material is responsible for limiting the output ratio of the primary colored material in local areas and ensuring that it does not block the light in the light path with high transmittance, so that the light can reach the deep layer or reflective substrate efficiently. Layer N-1 is defined as the reflection control layer, corresponding to the opaque primary color channel. The binary state "1" of voxels in this layer is mapped to the deposition of light-absorbing material related to the color presentation of the area, while "0" is mapped to the deposition of non-light-absorbing material related to light reflection, achieving the goal of finely adjusting the brightness and contrast of the upper layer material. For deep internal voxels with SDF values ​​greater than the total tinting depth D, the system defaults to filling with high-reflectivity non-light-absorbing material to maximize light energy utilization. This collaborative design of "internal reflection modulation + surface transparency color filtering" compensates for the light scattering loss of the translucent material from a physical mechanism perspective. A schematic diagram of the coloring of multilayer materials is shown below. Figure 3 As shown.

[0087] This application employs a fixed multilayer material stacking order, arranging the output order of the primary color channels based on differences in their transmittance and apparent brightness properties. Primary color channels with higher transmittance and apparent brightness are positioned at the outermost edge of the optical path, with other channels sequentially pushed inwards. The purpose of this layer sequence configuration is to ensure that incident light passes through a low optical density layer before entering the highly absorbent dark functional layer, thereby reducing initial absorption loss, increasing effective return light flux, and maintaining the overall reflective brightness of the printed surface. In the multilayer media structure, this fixed layer sequence defines a clear distribution of spectral absorption and scattering weights, enabling a predictable subtractive color mixing process as light reaches the underlying highly reflective substrate and is scattered multiple times before returning to the observation direction. At the computational engineering level, the fixed functional layer stacking order provides stable and consistent boundary conditions for multilayer color prediction models based on physical optics theory, making it possible to construct multilayer color lookup tables at the discrete voxel level. By constraining the spatial topological relationships between functional layers, this application can statistically reflect the interface reflection and ink layer light penetration behavior under a specific layer sequence, thereby improving the stability and accuracy of the computational fitting from the voxel state of digital materials to the final output color prediction.

[0088] Step S3: In each logic layer, perform 3D halftone screening independently to generate the initial voxel binary distribution of each primary color channel.

[0089] Specifically, for the color voxel generation process in multi-material 3D printing, a channel-independent error diffusion mechanism with logical layer constraints is adopted. Halftone screening is performed on the functional layers of different primary color channels to generate the initial voxel binary distribution of each primary color channel. The core of halftone screening is to convert continuous tone color values ​​into discrete voxel binary states while ensuring color reproduction accuracy and surface detail clarity.

[0090] Based on this, step S3 includes:

[0091] Generate a geometrically adaptive voxel traversal sequence, which is constructed based on the local geometric properties and spatial topological relationships of the model surface;

[0092] Along the voxel traversal sequence, three-dimensional error diffusion processing is performed independently in each logic layer to generate the initial voxel binary distribution of each primary color channel.

[0093] Generating geometrically adaptive voxel traversal sequences includes:

[0094] Divide the object to be printed into multiple slice layers along the construction direction;

[0095] Identify the connected components within the current slice layer;

[0096] For each surface connected component in the current slice layer, a traversal path is generated using a serpentine traversal strategy or a directed traversal strategy, depending on whether the connected component is newly appearing or about to disappear. The directed traversal strategy includes: determining the traversal starting point and priority traversal direction based on the component of the voxel surface normal in the slice normal direction, and selecting the next voxel during the traversal process based on directional consistency and the distance relationship between the voxel and the empty region.

[0097] The three-dimensional error propagation process is performed independently within each logic layer, including:

[0098] For the current voxel in the traversal sequence, a local tangent plane and its two-dimensional coordinate system are dynamically constructed based on its surface normal and local traversal direction;

[0099] Map the predefined two-dimensional error diffusion filter to the local tangent plane;

[0100] Within the 3D neighborhood of the current voxel, select a neighboring voxel that matches the filter weight point at the tangent plane projection position and is located after the current voxel in the traversal sequence. Distribute the error generated by the quantization of the current voxel to the neighboring voxel according to the filter weight ratio.

[0101] Specifically, step S3 includes the following sub-steps:

[0102] First, color sampling is required. The sampling operation is only performed on the visible surface voxels of the model. The continuous tone color value corresponding to the surface voxel is obtained as the color source data. For the other internal voxels, their target color value is not directly sampled from the surface. Instead, based on the preset spatial relationship of the voxel in three-dimensional space, the nearest surface voxel is determined, and the continuous tone color value of the surface voxel is inherited, or the projected color value of the surface voxel in the voxel normal or traversal direction is inherited. This avoids repeatedly performing color sampling for a large number of internal voxels and greatly reduces the computational overhead.

[0103] Then, a geometrically adaptive voxel traversal sequence is generated. This sequence is constructed based on the local geometric properties and spatial topological relationships of the model surface, ensuring that the traversal order conforms to the surface shape of the model and avoiding structural artifacts. The specific generation process is as follows:

[0104] The object to be printed is divided into multiple slices in the form of voxels along the construction direction. The slices are parallel to each other and orthogonal to the construction direction, which facilitates streaming data processing.

[0105] For the set of voxels in the current slice layer, the system identifies the connected components in the slice layer based on the spatial adjacency relationship between the voxels. Each connected component consists of voxels that are adjacent to each other in the slice layer and belong to the same target object.

[0106] For each surface-connected component in the current slice layer, a traversal path is generated using either a serpentine traversal strategy or a directed traversal strategy, depending on whether the connected component is newly appearing or about to disappear:

[0107] If the connected component is a newly appearing or soon-to-disappear component, a snake traversal strategy is adopted to simplify the traversal process and improve processing efficiency.

[0108] If the connected components are persistent, a directed traversal strategy is adopted: the traversal starting point and priority traversal direction are determined based on the component of the voxel surface normal in the slice normal direction. When the component of the starting voxel surface normal in the slice normal direction is negative, the voxel farthest from the empty region in the tangent plane is selected as the starting point; when the component is positive, the voxel closest to the empty region is selected as the starting point. During the traversal, the next voxel is selected based on directional consistency and the distance relationship between the voxel and the empty region to ensure the continuity and geometric consistency of the traversal path. If there is no candidate voxel that meets the conditions under the current traversal direction, the traversal direction can be reversed and the neighborhood search can be re-executed; if a valid candidate voxel is still not obtained after reversing, a new traversal starting point is selected from the untraversed surface voxels in the current slice, and the traversal continues.

[0109] Finally, following the voxel traversal sequence, three-dimensional error diffusion processing is performed independently within each logic layer to generate the initial voxel binary distribution for each primary color channel. The purpose of error diffusion processing is to evenly distribute the errors generated during voxel quantization to the unprocessed voxels, reducing halftone artifacts. The specific process is as follows:

[0110] After determining the starting voxel, an initial traversal direction is specified for the current traversal process. If the voxel has received errors from adjacent slices during the error propagation process of previous slices, the initial traversal direction is set to the reverse of the traversal direction used when the most recent error propagated to the voxel; otherwise, a preset default principal direction (such as clockwise or counterclockwise around a line parallel to the z-axis) is used to initialize the traversal direction. This direction serves as a global constraint for the current connected components, ensuring that the traversal process maintains a consistent clockwise or counterclockwise progression pattern, thereby achieving a uniform error distribution effect for 2D serpentine scanning on the surface voxel structure and avoiding directional stripe artifacts and local error accumulation.

[0111] During the traversal, for the current voxel, unvisited neighboring voxels are searched in its local neighborhood (e.g., a 3×3×3 neighborhood) as candidate voxels. Candidate voxels must satisfy the condition that their positional orientation relative to the current voxel is consistent with the predetermined traversal direction. Among candidate voxels that satisfy the orientation consistency, further selection is based on their distance from empty regions: when the component of the current voxel's surface normal in the slice normal direction is negative, the candidate voxel farthest from the empty region is selected first; when the component is positive, the candidate voxel closest to the empty region is selected first to ensure path continuity.

[0112] If no candidate voxels satisfy the conditions are found in the current traversal direction, the traversal direction is reversed and the neighborhood search is re-executed. If no valid candidate voxels are found after reversing, the traversal path is considered to have reached its end. The system then selects a new traversal starting point from the untraversed surface voxels in the current slice and sets its initial traversal direction to the reverse of the traversal direction used when propagating errors to that voxel, in order to continue traversing the remaining voxels in the current slice. Once all voxels in the current slice have been traversed, the traversal process switches to the next slice and continues according to the same rules, thus forming a continuous and geometrically adaptive traversal sequence in the entire 3D voxel space.

[0113] During the generation of the aforementioned traversal sequence and the execution of binarization error propagation, for the current voxel in the traversal sequence, a local tangent plane and its two-dimensional coordinate system are dynamically constructed based on its surface normal and local traversal direction. A predefined two-dimensional error propagation filter is then mapped to this local tangent plane. Within the three-dimensional neighborhood of the current voxel, a symmetric nearest-point matching method is used to select neighboring voxels that best match the filter weights at the tangent plane projection position, are located after the current voxel in the traversal sequence, and have not yet been processed. The error generated by the quantization of the current voxel is then distributed to these neighboring voxels according to the filter weight ratio. This error propagation process is performed independently in different layers, and the generated errors are not propagated between layers.

[0114] This embodiment adapts the error diffusion technology in two-dimensional image processing to three-dimensional geometric space. Under the premise of strictly preventing error backflow, it realizes the smooth and directional diffusion of quantization error on complex surfaces, so that high-fidelity, low-artifact full-color 3D printing results can still be obtained when using semi-transparent printing materials.

[0115] Step S4: Based on the depth optimization method, the depth unit formed along the surface voxel normal direction is used as the basic evaluation unit. Under the premise of keeping the total volume coverage of each primary color channel material unchanged, the initial voxel binary distribution is iteratively optimized through local voxel exchange operation until the preset conditions are met, and the optimized voxel distribution is obtained for three-dimensional color printing.

[0116] Specifically, after completing the multi-channel 3D error diffusion processing, based on the physical optics model, using depth units formed along the surface voxel normal direction as the basic evaluation unit, and while maintaining the total volume coverage of each primary color channel material, the initial voxel binary distribution is iteratively optimized through local voxel exchange operations until preset conditions are met, resulting in an optimized voxel distribution for 3D color printing. The core of this step is to compensate for color deviations caused by volumetric optical effects and improve color reproduction accuracy by controlling the stacking structure of multilayer materials.

[0117] Based on this, and using a depth optimization method, the basic evaluation unit is the depth element formed along the surface voxel normal direction, including:

[0118] For any voxel location on the visible surface of an object, a combination of voxels extending along the surface normal of that voxel and covering the multiple logic layers is defined as a depth unit. The depth unit is used to fully describe the material stacking structure of the surface point along the normal direction.

[0119] Iterative optimization of the initial binary voxel distribution through local voxel swap operations includes:

[0120] For each depth cell, its predicted color value is calculated based on the depth optimization method;

[0121] Calculate the color difference between the predicted color value and the target color value, and use it as the color error of the surface voxel corresponding to the depth unit;

[0122] With the goal of minimizing the sum of color errors of all surface voxels, the voxel distribution is iteratively optimized through local voxel exchange operations.

[0123] The predicted color values ​​for depth cells calculated based on the physical optics model include:

[0124] The Kubelka-Monk model was used to perform optical calculations on the multilayer material stacked structure within the depth unit to obtain the comprehensive spectral reflectance.

[0125] The comprehensive spectral reflectance is converted into the International Commission on Illumination (ICI) standard color space value under standard light source conditions and mapped to the ICI uniform color space to obtain the predicted color value;

[0126] The Kubelka-Monk model calculation is accelerated by a pre-built multi-layer color lookup table.

[0127] Local voxel exchange operations include:

[0128] Within the same primary color channel layer, with the current voxel as the center, search for candidate voxels located in the same slice and whose binary state is opposite to that of the current voxel within its neighborhood;

[0129] Swap the binary states of the current voxel and the candidate voxel on the primary color channel layer;

[0130] The swap operation is only accepted when the local color difference is reduced after the swap, and the total number of voxels in the primary color channel layer with a state of 1 remains unchanged before and after the swap.

[0131] Specifically, first define the depth unit, such as Figure 4 As shown, for any voxel position located on the visible surface of the object Define a direction along the normal direction of the voxel surface. The voxel combination that extends and covers the multiple logic layers serves as a depth unit. The depth unit is used to fully describe the material stacking structure of surface points along the normal direction and is the basic unit for color evaluation and optimization.

[0132] Secondly, the predicted color value of the depth unit is calculated based on the physical optics model. The physical optics model is then used to perform optical calculations on the multilayer material stacked within the depth unit to obtain the comprehensive spectral reflectance. The physical optics model can accurately describe the light absorption and scattering characteristics of translucent materials. Combined with the pre-measured reflectance R and transmittance T of each basic material, the absorption coefficient Ki and scattering coefficient Si of the material are calculated, thereby simulating the multiple scattering and absorption process of light in the multilayer material. Since each material layer has a geometrically fixed thickness close to that of a single voxel, and the material type values ​​are limited, the above multilayer optical synthesis process can be accelerated by a pre-constructed multilayer color lookup table, significantly improving color prediction efficiency while ensuring physical consistency. The physical optics model can be a multilayer optical calculation model based on radiative transfer theory, including but not limited to the Kubelka–Munk model, the Four-Flux Model, the Monte Carlo method, or other equivalent models capable of describing the optical behavior of multilayer scattering media.

[0133] The comprehensive spectral reflectance is converted to CIE XYZ values ​​under standard light source conditions and mapped to the CIELAB space to obtain the predicted color value. The CIELAB color space accurately reflects the differences in human color perception, providing a reliable basis for color difference calculation. It compares the predicted color value with the target color given by the continuous tone color volume. The color difference between them is dynamically adjusted to adjust the material distribution within the depth unit.

[0134] The color error is calculated and the optimization target is determined. The color difference between the predicted color value and the target color value is calculated and used as the color error of the surface voxel corresponding to the depth unit. ,in To predict color values, Let be the target color value. The overall objective of deep local optimization is to minimize the sum of color errors of all surface voxels, and its global energy function is: in, , This represents the set of all surface voxels involved in coloring. Iterative optimization stops when no effective exchange occurs or the preset number of iterations is reached during a single scan. By locally rearranging the voxel distribution, the energy function gradually decreases while maintaining the total coverage of each material. This suppresses unfavorable overprint combinations in space, significantly improving the color accuracy and visual consistency of multi-material 3D printing results.

[0135] The local voxel exchange operation and iterative optimization aim to minimize the sum of color errors of all surface voxels. The voxel distribution is iteratively optimized through local voxel exchange operations. The specific operation process is as follows:

[0136] Within the same primary color channel layer, with the current voxel P A Centered on the current voxel, search within its neighborhood for candidate voxels P that are located in the same slice and whose binary state is opposite to the current voxel. B Candidate voxel P B The conditions must be met: the voxel must not have been processed yet, it must be located in the same slice as the current voxel, and it must belong to the same primary color channel layer. This ensures that the swap operation does not cross layers and does not affect the voxel distribution of other channels.

[0137] Candidate voxel P B The binarization result on the material channel needs to be compared with the current voxel P. A Conversely, the current voxel P A With the candidate voxel P B The binary states on the primary color channel layer are swapped; this swapping operation is only accepted if the local color difference is reduced after the swap. That is, it satisfies... ,in This is the sum of the color errors before the exchange. This is the sum of the color errors after the exchange. For small positive numbers used to suppress numerical oscillations, , This represents the error recalculated using a lookup table after local voxel material exchange. The total number of voxels in state 1 in the primary color channel layer remains unchanged before and after the exchange, ensuring that the total volume coverage of each primary color channel material remains constant and does not disrupt the material distribution ratio determined during the error diffusion stage.

[0138] Repeat the above local voxel exchange operation until the preset optimization conditions are met. The preset conditions can be set as follows: no effective exchange occurs during a certain scan, or the number of iterations reaches a preset threshold. During the iterative optimization process, the voxel access order generated in the error diffusion stage is fully inherited. By maintaining the consistency of the traversal order, it is possible to avoid introducing new periodic structures or directional artifacts during the optimization process, thereby avoiding the disruption of the high-frequency dominant, non-periodic randomized spatial distribution characteristics formed in the error diffusion stage.

[0139] Compared with the prior art, the present invention has the following significant advantages and beneficial effects:

[0140] This invention constructs a primary color material system based on the transparency properties of materials, and combines a directed distance field to achieve precise layering of surface voxels, corresponding each primary color channel to a logical layer. At the same time, it introduces a physical optics model to accurately describe the volumetric optical effects of multi-layered semi-transparent materials. Through deep iterative optimization to compensate for color deviation, it effectively solves the problems of low color saturation, narrow color gamut, and blurred surface details in traditional methods, and significantly improves the color reproduction accuracy and visual consistency of 3D color printing.

[0141] This invention employs independent halftone screening for each channel, ensuring that the error propagation process of each logic layer does not interfere with each other. At the same time, it reduces invalid calculations through geometrically adaptive voxel traversal sequences. In the deep optimization stage, local voxel exchange operations are used, which only adjust the local neighborhood without requiring global recalculation. Combined with multi-layer color lookup tables, optical calculations are accelerated. While ensuring the optimization effect, this invention achieves efficient processing of large-scale voxel data, meeting the engineering requirements of high-resolution 3D printing.

[0142] This invention is applicable to color printing materials with various transparency combinations. The number of primary color channels, layer parameters, and optimization strategies can be flexibly adjusted according to the characteristics of the printing equipment and the target color requirements to adapt to different 3D printing scenarios. It is especially suitable for high-fidelity 3D color printing tasks with complex curved surfaces and has broad application prospects.

[0143] By using geometrically adaptive voxel traversal sequences and channel-specific error diffusion processing, the striped or checkerboard structural artifacts caused by inconsistent screen orientation in complex curved areas by traditional methods are effectively suppressed. The randomness of voxel distribution is maintained during the depth optimization stage, which further avoids the introduction of new artifacts during the optimization process and improves the surface quality of the printed product.

[0144] It should be noted that the above description describes some embodiments of the present invention. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps described in the claims may be performed in a different order than that shown in the above embodiments and still achieve the desired results. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0145] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of the invention (including the claims) is limited to these examples; within the framework of the invention, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of the embodiments of the invention as described above, which are not provided in the details for the sake of brevity.

[0146] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0147] While specific details have been set forth to describe exemplary embodiments of the invention, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details or with variations thereof. Therefore, these descriptions should be considered illustrative rather than restrictive. Although the invention has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art based on the foregoing description.

[0148] The embodiments of this invention are intended to cover all such substitutions, modifications, and variations falling within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the embodiments of this invention should be included within the protection scope of this invention.

Claims

1. A depth-optimized three-dimensional color control method based on color printing materials, characterized in that, include: S1, Construct a primary color material system, wherein the primary color material system is classified according to the transparency properties of the printing material and assigned to the corresponding primary color channels; S2, the 3D model to be printed is voxelized, and the voxels on the surface of the object are divided into multiple logical layers corresponding to the number of primary color channels inward along the normal direction based on the directed distance field. Each logical layer corresponds to one primary color channel. S3, within each logic layer, independently performs three-dimensional halftone screening to generate the initial voxel binary distribution of each primary color channel; S4. Based on the depth optimization method, the depth unit formed along the surface voxel normal direction is used as the basic evaluation unit. Under the premise of keeping the total volume coverage of each primary color channel material unchanged, the initial voxel binary distribution is iteratively optimized through local voxel exchange operation until the preset conditions are met, and the optimized voxel distribution is obtained for three-dimensional color printing.

2. The depth-optimized three-dimensional color control method based on color printing materials according to claim 1, characterized in that, The method of dividing the object surface voxels into multiple logical layers corresponding to the number of primary color channels inward along the normal direction based on the directed distance field includes: Calculate the directed distance from each voxel in the voxel mesh to the object surface; The layer thickness parameter is determined based on the physical resolution of the printing device, and the layer thickness parameter is equal to the voxel physical size of the printing system in the direction of minimum resolution. Based on the layer thickness parameter, a set of equidistant distance thresholds are constructed, and a series of equidistant shells are extracted in the directed distance field using the distance thresholds, dividing the object surface voxels into geometric layers arranged in order corresponding to the number of primary color channels. The geometric layers are sequentially mapped to functional layers, and each voxel is assigned a layer index, establishing a mapping relationship from three-dimensional geometric space to multi-layer shading logic.

3. The depth-optimized three-dimensional color control method based on color printing materials according to claim 2, characterized in that, The primary color material system includes opaque materials, translucent materials, and transparent materials, and the functional layer includes a color modulation layer and a reflective substrate layer; Among them, the first to N-2 layers from the outside to the inside are color modulation layers, corresponding to semi-transparent or transparent base color materials; the N-1 layer is a reflective base layer, corresponding to non-transparent base color materials.

4. The depth-optimized three-dimensional color control method based on color printing materials according to claim 1, characterized in that, Step S3 includes: Generate a geometrically adaptive voxel traversal sequence, which is constructed based on the local geometric properties and spatial topological relationships of the model surface; Along the voxel traversal sequence, three-dimensional error diffusion processing is performed independently in each logic layer to generate the initial voxel binary distribution of each primary color channel.

5. The depth-optimized three-dimensional color control method based on color printing materials according to claim 4, characterized in that, The generated geometry-adaptive voxel traversal sequence includes: Divide the object to be printed into multiple slice layers along the construction direction; Identify the connected components within the current slice layer; For each surface connected component in the current slice layer, a traversal path is generated using a serpentine traversal strategy or a directed traversal strategy, depending on whether the connected component is newly appearing or about to disappear. The directed traversal strategy includes: determining the traversal starting point and priority traversal direction based on the component of the voxel surface normal in the slice normal direction, and selecting the next voxel during the traversal process based on directional consistency and the distance relationship between the voxel and the empty region.

6. The depth-optimized three-dimensional color control method based on color printing materials according to claim 4, characterized in that, The independent execution of three-dimensional error diffusion processing within each logic layer includes: For the current voxel in the traversal sequence, a local tangent plane and its two-dimensional coordinate system are dynamically constructed based on its surface normal and local traversal direction; Map the predefined two-dimensional error diffusion filter to the local tangent plane; Within the 3D neighborhood of the current voxel, select a neighboring voxel whose projection position on the tangent plane matches the filter weight point and whose position is after the current voxel in the traversal sequence. Distribute the error generated by the quantization of the current voxel to the neighboring voxel according to the filter weight ratio.

7. The depth-optimized three-dimensional color control method based on color printing materials according to claim 1, characterized in that, The depth optimization method, which uses depth units formed along the surface voxel normal direction as the basic evaluation unit, includes: For any voxel location on the visible surface of an object, a combination of voxels extending along the surface normal of that voxel and covering the multiple logic layers is defined as a depth unit. The depth unit is used to fully describe the material stacking structure of the surface point along the normal direction.

8. The depth-optimized three-dimensional color control method based on color printing materials according to claim 1 or 7, characterized in that, The iterative optimization of the initial binary voxel distribution through local voxel exchange operations includes: For each depth cell, the predicted color value is calculated based on the depth optimization method; Calculate the color difference between the predicted color value and the target color value, and use it as the color error of the surface voxel corresponding to the depth unit; With the goal of minimizing the sum of color errors of all surface voxels, the voxel distribution is iteratively optimized through local voxel exchange operations.

9. The depth-optimized three-dimensional color control method based on color printing materials according to claim 8, characterized in that, The calculation of the predicted color value based on the depth optimization method includes: Optical calculations were performed on the multi-layer material stacked structure within the depth unit using a physical optics model to obtain the comprehensive spectral reflectance. The comprehensive spectral reflectance is converted into the International Commission on Illumination (ICI) standard color space value under standard light source conditions and mapped to the ICI uniform color space to obtain the predicted color value; The optical calculation process is accelerated by a pre-built multi-layer color lookup table.

10. The depth-optimized three-dimensional color control method based on color printing materials according to claim 8, characterized in that, The local voxel exchange operation includes: Within the same primary color channel layer, with the current voxel as the center, search for candidate voxels located in the same slice and whose binary state is opposite to that of the current voxel within its neighborhood; Swap the binary states of the current voxel and the candidate voxel on the primary color channel layer; The swap operation is only accepted when the local color difference is reduced after the swap, and the total number of voxels in the primary color channel layer with a state of 1 remains unchanged before and after the swap.

Images