Great research starts with great data.

Learn More
More >
Patent Analysis of

METHOD AND SYSTEM FOR PRINTING 3D OBJECT

Updated Time 15 March 2019

Patent Registration Data

Publication Number

US20180162057A1

Application Number

US15/882690

Application Date

29 January 2018

Publication Date

14 June 2018

Current Assignee

ZHUHAI SEINE TECHNOLOGY CO., LTD.

Original Assignee (Applicant)

ZHUHAI SEINE TECHNOLOGY CO., LTD.

International Classification

B29C64/393,B29C64/124,G06F17/50,B41F15/08

Cooperative Classification

B29C64/393,B29C64/124,G06F17/50,B41F15/0881,B33Y10/00

Inventor

CHEN, XIAOKUN,JIANG, WEI,CHEN, WEI,ZHOU, YI

Abstract

The present invention provides a method and system for printing a 3D object. The method for printing a 3D object includes the following steps: a. performing conversion based on layered image data DM of a target object to obtain printing data PMN, where DM consists of a value set XMN of NM pixel points, M represents the number of layers of a layered image, and N represents the number of pixel points; b. performing layer-by-layer printing based on the printing data PMN, where when PMN is 0, a first printing material is used for printing, and when the printing data PMN is 1, a second printing material is used for printing; and c. superposing results of the layer-by-layer printing in step b to form a 3D object.

Read more

Claims

1. A method for printing a 3D object, comprising the following steps: a. performing conversion based on layered image data DM of a target object to obtain printing data PMN, wherein DM consists of a value set XMN of NM pixel points, M represents the number of layers of a layered image, and N represents the number of pixel points; b. performing layer-by-layer printing based on the printing data PMN, wherein when the printing data PMN is 0, a first printing material is used for printing, and when the printing data PMN is 1, a second printing material is used for printing; and c. superposing results of the layer-by-layer printing in step b to form a 3D object.

2. The printing method according to claim 1, wherein in step b, when the printing data PMN is 0, correction is performed on the printing data PMN to obtain corrected printing data PMN′, and when the corrected printing data PMN′ is identified, the first printing material is used for printing.

3. The printing method according to claim 2, wherein in step b, an association is established between the corrected printing data PMN′ and a printing startup instruction, and when the corrected printing data PMN′ is identified, the first printing material is used for printing.

4. The printing method according to claim 3, wherein lamination and unit-division processing is performed on the entire target object to obtain the layered image data DM.

5. The printing method according to claim 4, wherein in step a, half-tone conversion is performed on the layered image data DM to obtain the printing data PMN.

6. The printing method according to claim 5, wherein the layered image data DM is 8-bit data, and values of XMN corresponding to DM are between 0 and 255.

7. The printing method according to claim 6, wherein the printing data PMN is obtained in the following manner: a1. dividing the value set XMN by 255 to obtain a corresponding fitted value set XMN′, wherein the fitted value set XMN′ is between 0 and 1, and the fitted value set XMN′ consists of x11′, x12′, x21′, x22′, . . . , and xMN′, and a2. setting a threshold S, and comparing the fitted value set XMN′ with the threshold S, wherein if xMN′ is less than S, converted pMN is 0; or if xMN′ is greater than S, converted pMN is 1, and the printing data PMN that is correspondingly formed consists of a series of pMN.

8. The printing method according to claim 6, wherein the printing data PMN is obtained in the following manner: a1′. dividing the value set XMN by 255 to obtain a corresponding fitted value set XMN′, wherein the fitted value set XMN′ sequentially consists of xM1′, xM2′, xM3′, . . . , and xMN′ according to an order of neighboring relationships of N pixel points of an Mth-layer image; a2′. comparing a threshold S with xM1′, wherein if xM1′ is less than S, pM1 is 0; or if xM1′ is greater than S, pM1 is 1; a3′. calculating a difference EMN by subtracting pMN from xMN′, and calculating a sum of xM(N+1)′ and EMN to obtain xM(N+1)″; a4′. comparing xM(N+1)″ with the threshold S, wherein if xM(N+1)″ is less than S, pM(N+1) is 0; or if xM(N+1)″ is greater than S, pM(N+1) is 1; and a5′. repeating steps a3′ and a4′ until all xMN′ are converted into pMN, wherein the printing data PMN that is correspondingly formed consists of a series of pMN.

9. The printing method according to claim 7, wherein S is any one of the following values: 0.4; 0.5; 0.55; 0.6; or 0.65.

10. The printing method according to claim 8, wherein S is any one of the following values: 0.4; 0.5; 0.55; 0.6; or 0.65.

11. The printing method according to claim 6, wherein the printing data PMN is obtained in the following manner: a1″. dividing the value set XMN by 255 to obtain a corresponding fitted value set XMN′, and sorting the fitted value set XMN′ according to N pixel points of an Mth-layer image, wherein the fitted value set XMN′ sequentially consists of xM1′, xM2′, xM3′, xM4′, . . . , and xMN′; a2″. setting g thresholds and sorting the g thresholds to form a threshold set SMg′, wherein the threshold set SMg′ sequentially consists of sM1′, sM2′, sM3′, sM4′, . . . , and sMg′, and 0<g≤N; and a3″. based on a sequence of the threshold set SMg′, correspondingly comparing N values xMN′ of the Mth-layer image in the fitted value set XMN′ with the g thresholds sMg′ in the threshold set SMg′ one by one, wherein if xMN′ is less than sMg′, pMN is 0; or if xMN′ is greater than sMg′, pMN is 1, and the printing data PMN that is correspondingly formed consists of a series of pMN.

12. The printing method according to claim 11, wherein a value range of the threshold set SMg′ is any value between 0.4 and 0.65.

13. The printing method according to claim 5, wherein the layered image data DM is 16-bit data, and values of the value set XMN corresponding to DM are between 0 and 65535.

14. The printing method according to claim 1, wherein the first printing material is one of the following materials: a transparent material; a white material; or an off-white material.

15. The printing method according to claim 14, wherein the second printing material is one of the following materials: a combination of a cyan material, a magenta material, and a yellow material; or a combination of any two of a cyan material, a magenta material, and a yellow material.

16. The printing method according to claim 15, wherein the first printing material and the second printing material are one of the following materials: a light-curing material; or a temperature curing material.

17. The printing method according to claim 1, wherein in step c, each result of the layer-by-layer printing is cured and then superposed.

18. The printing method according to claim 17, wherein in step c, after being leveled, each result of the layer-by-layer printing is first cured and then superposed.

19. A system for printing a 3D object, comprising a data processor, a process controller, and a printhead, wherein the data processor converts layered image data DM of a target object to obtain printing data PMN, and performs correction on data being 0 in the printing data PMN, to obtain corrected printing data PMN; and the process controller controls, based on the printing data PMN and the corrected printing data PMN′, the printhead to perform layer-by-layer printing and superposition, to form a 3D object, wherein based on the corrected printing data PMN′, a first printing material is used for printing, and based on the printing data PMN, a second printing material is used for printing.

20. The system according to claim 19, further comprising a leveling apparatus, wherein the leveling apparatus is configured to perform leveling on each result of the layer-by-layer printing.

21. The system according to claim 20, further comprising a curing apparatus, wherein the curing apparatus is configured to perform curing on each result of the layer-by-layer printing.

Read more

Claim Tree

  • 1
    1. A method for printing a 3D object, comprising
    • the following steps: a. performing conversion based on layered image data DM of a target object to obtain printing data PMN, wherein DM consists of a value set XMN of NM pixel points, M represents the number of layers of a layered image, and N represents the number of pixel points
    • b. performing layer-by-layer printing based on the printing data PMN, wherein when the printing data PMN is 0, a first printing material is used for printing, and when the printing data PMN is 1, a second printing material is used for printing
    • and c. superposing results of the layer-by-layer printing in step b to form a 3D object.
    • 2. The printing method according to claim 1, wherein
      • in step b, when the printing data PMN is 0, correction is performed on the printing data PMN to obtain corrected printing data PMN′, and when the corrected printing data PMN′ is identified, the first printing material is used for printing.
    • 14. The printing method according to claim 1, wherein
      • the first printing material is one of the following materials: a transparent material; a white material; or an off-white material.
    • 17. The printing method according to claim 1, wherein
      • in step c, each result of the layer-by-layer printing is cured and then superposed.
  • 19
    19. A system for printing a 3D object, comprising
    • a data processor, a process controller, and a printhead, wherein the data processor converts layered image data DM of a target object to obtain printing data PMN, and performs correction on data being 0 in the printing data PMN, to obtain corrected printing data PMN
    • and the process controller controls, based on the printing data PMN and the corrected printing data PMN′, the printhead to perform layer-by-layer printing and superposition, to form a 3D object, wherein based on the corrected printing data PMN′, a first printing material is used for printing, and based on the printing data PMN, a second printing material is used for printing.
    • 20. The system according to claim 19, further comprising
      • a leveling apparatus, wherein the leveling apparatus is configured to perform leveling on each result of the layer-by-layer printing.
See all 2 independent claims

Description

BACKGROUND

Technical Field

The present invention relates to a rapid prototyping technology of a 3D object, in particular, to a technology in which an inkjet printhead is used to perform layer-by-layer additive manufacturing of a 3D object, and specially, to a method and system for printing a 3D object.

Related Art

A rapid prototyping technology is also referred to as a rapid prototype manufacturing technology. Under control of a computer and based on the discrete/stacking principle, a physical shape of an object is converted into a three-dimensional digital model by using modeling software or a three-dimensional scanner, and the model is used to generate an STL file. Lamination software is used to discrete the model on Z-axis, to form a series of thin sheets having a same thickness or having different thicknesses, then a fused deposition modeling technology (FDM technology), a stereo lithography appearance technology (SLA technology), a selective laser sintering technology (SLS technology), a laminated object manufacturing technology (LOM technology), or the like is utilized to process and stack the series of thin sheets layer by layer, and finally, post-processing is performed to obtain a 3D image.

Compared with the conventional SLA, SLS, and LOM technologies, for a multi jet printing technology (MJP technology), according to the operating principle of an ink-jet printer, liquid (a building material and/or a manufacturing material) in a nozzle chamber instantly forms droplets under excitation of a digital signal, and the droplets are ejected from the nozzle at a speed and frequency. The droplets are cured and formed layer by layer according to a specified path, so that a 3D image is finally obtained. Because an expensive laser system is not required when the MJP is used, a device is cheap, and operation and maintenance costs are also very low. Compared with the FDM technology, work can be performed at a lower temperature when the MJP technology is used. Moreover, the MJP technology has such advantages as simple operation, high forming speed, high applicability to multiple materials, high precision of a formed member, and capability of being used in an office environment. Therefore, the MJP technology is one of current research focuses of the rapid prototyping technology.

Objects that are manufactured by using the MJP technology may be classified into two categories according to surface colors: one category is monochromatic objects, and the other category is colorful objects (including at least two colors). As demands on material culture increase, people are no more satisfied with single visual sense, and they expect to manufacture any colorful object having a personalized design by fully utilizing the advantages of the MJP technology.

An existing method for printing a 3D object on the market is that: different areas of a model file are manually colored, lamination processing is performed on the colored model by using lamination software, data information generated after the lamination processing is sent to a printer driving controller, the driving controller controls a printhead to eject ink along a predetermined path, layer-by-layer superposition is performed after curing, and finally a colorful object is printed out.

A disadvantage in the prior art is that the technology still remains stuck at a technical level of color selection for a module. In this case, it needs to be manually determined to add a color to a module. Such manual selection consumes time and labor power, and expression of color transition is poor. Moreover, different people have different abilities to discriminate between colors, and consequently, there may be color difference between a printed object and an object that is actually needed.

SUMMARY

For the disadvantage in the prior art, according to one aspect of the present invention, a method for printing a 3D object is provided, including the following steps:

a. performing conversion based on layered image data DM of a target object to obtain printing data PMN, wherein DM consists of a value set XMN of NM pixel points, M represents the number of layers of a layered image, and N represents the number of pixel points;

b. performing layer-by-layer printing based on the printing data PMN, wherein when the printing data PMN is 0, a first printing material is used for printing; and

c. superposing results of the layer-by-layer printing in step b to form a 3D object.

Further, in step b, when the printing data PMN is 0, correction is performed on the printing data PMN to obtain corrected printing data PMN′, and when the corrected printing data PMN′ is identified, the first printing material is used for printing.

Further, in step b, an association is established between the corrected printing data PMN′ and a printing startup instruction, and when the corrected printing data PMN′ is identified, the first printing material is used for printing.

Further, lamination and unit-division processing is performed on the entire target object to obtain the layered image data DM.

Further, in step a, half-tone conversion is performed on the layered image data DM to obtain the printing data PMN.

Further, the layered image data DM is 8-bit data, and values of XMN corresponding to DM are between 0 and 255.

Further, the printing data PMN is obtained in the following manner:

a1. dividing the value set XMN by 255 to obtain a corresponding fitted value set XMN′, where the fitted value set XMN′ is between 0 and 1, and the fitted value set XMN′ consists of x11′, x12′, x21′, x22′, . . . , and xMN′; and

a2. setting a threshold S, and comparing the fitted value set XMN′ with the threshold S, where if xMN′ is less than S, converted pMN is 0; or if xMN′ is greater than S, converted pMN is 1, and the printing data PMN that is correspondingly formed consists of a series of pMN.

Further, the printing data PMN is obtained in the following manner:

a1′. dividing the value set XMN by 255 to obtain a corresponding fitted value set XMN′, where the fitted value set XMN′ sequentially consists of xM1′, xM2′, xM3′, . . . , and xMN′ according to an order of neighboring relationships of N pixel points of an Mth-layer image;

a2′. comparing a threshold S with xM1′, where if xM1′ is less than S, pM1 is 0; or if x11′ is greater than S, pM1 is 1;

a3′. calculating a difference EMN by subtracting pMN from xMN′, and calculating a sum of xM(N+1)′ and EMN to obtain xM(N+1)″;

a4′. comparing xM(N+1)″ with the threshold S, where if xM(N+1)″ is less than S, pM(N+1) is 0; or if xM(N+1)″ is greater than S, pM(N+1) is 1; and

a5′. repeating steps a3′ and a4′ until all xMN are converted into pMN, where the printing data PMN that is correspondingly formed consists of a series of pMN.

Further, S is any one of the following values: 0.4; 0.5; 0.55; 0.6; or 0.65.

Further, the printing data PMN is obtained in the following manner:

a1″. dividing the value set XMN by 255 to obtain a corresponding fitted value set XMN′, and sorting the fitted value set XMN′ according to N pixel points of an Mth-layer image, where the fitted value set XMN′ sequentially consists of xM1′, xM2′, xM3′, xM4′, . . . , and xMN′;

a2″. setting g thresholds and sorting the g thresholds to form a threshold set SMg′, wherein the threshold set SMg′ sequentially consists of sM1′, sM2′, sM3′, sM4′, . . . , and sMg′, and 0<g≤N; and

a3″. based on a sequence of the threshold set SMg′, correspondingly comparing N values of the Mth-layer image in the fitted value set XMN′ with the g thresholds sMg′ in the threshold set SMg′ one by one, where if xMN′ is less than sMg′, pMN is 0; or if xMN′ is greater than sMg′, pMN is 1, and the printing data PMN that is correspondingly formed consists of a series of pMN.

Further, a value range of the threshold set SMg′ is any value between 0.4 and 0.65.

Further, the layered image data DM is 16-bit data, and values of XMN corresponding to DM are between 0 and 65535.

Further, the first printing material is one of the following materials:

a transparent material;

a white material; or

an off-white material.

Further, in step b, when PMN is 1, a second printing material is used for printing.

Further, the second printing material is one of the following materials:

a combination of a cyan material, a magenta material, and a yellow material; or

a combination of any two of a cyan material, a magenta material, and a yellow material.

Further, the first printing material and the second printing material are one of the following materials:

a light-curing material; or

a temperature curing material.

Further, in step c, each result of the layer-by-layer printing is cured and then superposed.

Further, in step c, after being leveled, each result of the layer-by-layer printing is first cured and then superposed.

According to another aspect of the present invention, a system for printing a 3D object is provided, including a data processor, a process controller, and a printhead, wherein

the data processor converts layered image data DM of a target object to obtain printing data PMN, and performs correction on data being 0 in the printing data PMN, to obtain corrected printing data PMN′; and

the process controller controls, based on the printing data PMN and the corrected printing data PMN′, the printhead to perform layer-by-layer printing and superposition, to form a 3D object, wherein based on the corrected printing data PMN′, a first printing material is used for printing, and based on the printing data PMN, a second printing material is used for printing.

Further, the system further includes a leveling apparatus, where the leveling apparatus is configured to perform leveling on each result of the layer-by-layer printing.

Further, the system further includes a curing apparatus, where the curing apparatus is configured to perform curing on each result of the layer-by-layer printing.

According to the present invention, a layer-by-layer printing method is used. After lamination and unit-division is performed on an entire target object, layered image data DM is obtained, and half-tone conversion is performed on the layered image data DM to obtain printing data PMN. When PMN is 0, a first printing material is used for printing; or when PMN is 1, a second printing material is used for printing. A 3D object that is printed out by using the printing method of the present invention has stronger sense of layers. Especially when a colorful 3D object is printed, rich colors and natural transition between different colors can also be presented. Meanwhile, a method process of the present invention is simple, the automation degree is high, and a product is easily implemented.

BRIEF DESCRIPTION OF THE DRAWINGS

Other features, objectives, and advantages of the present invention will become more apparent with reference to detailed descriptions of non-limiting embodiments in the accompanying drawings:

FIG. 1 is a flowchart of a method for printing a 3D object according to a specific implementation of the present invention;

FIG. 2 is a flowchart of a method for obtaining printing data PMN by converting layered image data DM according to an embodiment of the present invention;

FIG. 3 is a flowchart of a method for obtaining printing data PMN by converting layered image data DM according to an embodiment of the present invention; and

FIG. 4 is a flowchart of a method for obtaining printing data PMN by converting layered image data DM according to an embodiment of the present invention.

DETAILED DESCRIPTION

FIG. 1 is a flowchart of a method for printing a 3D object according to a specific implementation of the present invention. Specifically, the method includes the following steps.

Step S101 is performed. Conversion is performed based on layered image data DM of a target object to obtain printing data PMN. DM consists of a value set XMN of NM pixel points. A person skilled in the art understands that, DM consists of image data of M layers, and includes d1, d2, d3, . . . , and dM; NM represents the number of pixel points included in the target object; XMN consists of values of the pixel points, and includes x11, x12, x21, x22, . . . , and xMN; and PMN consists of printing data values that are obtained after the conversion is performed on the values of the pixel points, and includes p11, p12, p21, p22, . . . , and pMN. Specifically, M represents the number of layers of a layered image, and N represents the number of the pixel points.

Further, the layered image data DM is data that is obtained after lamination and unit-division processing is performed on the target object. The layered image data DM corresponds to the value set XMN that consists of the values of the NM pixel points in the target object. The values of the NM pixel points include data of two aspects: spatial coordinates of the NM pixel points on a section on which the NM pixel points are located, and color values of the NM pixel points that are defined based on the section. Specifically, the spatial coordinates are obtained by performing lamination on the target object, and are defined by setting one or more coordinate axes in one or more directions. The color values may be color data based on a CMY color mode, or may be color data based on a CMYK color mode, or may be color data based on another color mode. The color values are defined based on a 2D printing technology.

Further, the printing data PMN is obtained after the conversion is performed on the layered image data DM. A person skilled in the art understands that there are many studies on a conversion method in the 2D printing field. According to the theory of image data conversion for 2D printing, in the 2D printing, an objective of performing the image data conversion is to implement obtaining, transmission, reproduction, and the like of an image with loss as little as possible. In a 3D printing technology, if conversion is only performed on specific layered image data, the basic principle of the conversion is similar to that of the 2D printing. However, when the layered image data is converted in the 3D printing, data of the layered image data in the 3D space, that is, information about spatial coordinates, also needs to be considered to better reproduce the target object.

Further, step S102 is performed. Layer-by-layer printing is performed based on the printing data PMN. A person skilled in the art understands that, during actual application, drive control data of most printheads is binary, and a corresponding value of the printing data PMN that is obtained after the conversion is performed is 0 or 1. Specifically, in an existing printing process, when the printing data PMN is 0, a printhead does not perform the printing. Only when the printing data PMN is 1, the printhead performs the printing. More specifically, in the present invention, a 3D object needs to be finally implemented. If a printing material is not ejected when the printing data PMN is 0, a pixel point at such position may be finally filled by a printing material of another pixel point, finally affecting a shape of the entire 3D object. Therefore, the printing data PMN that is obtained after the conversion is performed needs to be further processed, so that each pixel point is printed. That is, regardless of whether the printing data PMN is 0 or 1, the printhead is controlled to eject the printing material. More specifically, because drive control data of a common printhead is binary, a printing instruction received by the printhead is actually 0 or 1. The printhead does not perform the printing when the printing instruction is 0, that is, the printhead does not control a printing nozzle to eject the printing material. The printhead performs the printing when the printing instruction is 1, that is, the printhead controls the printing nozzle to eject the printing material. A process of performing further processing on the printing data PMN is to convert the printing data PMN whose value is 0, so that the printhead identifies the printing data PMN whose value is 0 as a printing instruction 1.

Further, based on performing step S102, when the printing data PMN is 0, step S103 is performed, and a first printing material is used for printing. When the printing data PMN is 1, step S104 is performed, and a second printing material is used for printing. That is, according to different values of the printing data PMN, the printhead is controlled to eject different printing materials. A specific process thereof is actually a process of converting the printing data PMN into different printing instructions. In an embodiment, the first printing material is a transparent material, a white material, or an off-white material, and the second printing material is a combination of a cyan material, a magenta material, and a yellow material. Specifically, when the second printing material is used for printing, a color that is used for printing needs to be selected according to a primary color of a section on which a pixel point to be printed is located. For example, the primary color of the section on which the pixel point to be printed is located is cyan, and the cyan material is used for printing. More specifically, determining of the primary color of the section is described in the foregoing step, and details are not described herein again. In a variation, the second printing material may alternatively be a combination of any two of a cyan material, a magenta material, and a yellow material, and another color material may also be added. This needs to be determined according to a color mode after half-tone conversion. A person skilled in the art understands that the first printing material and the second printing material may have multiple color combinations. Further, the first printing material and the second printing material are further light-curing materials or temperature curing materials.

In an embodiment, a method for performing further processing on the printing data PMN is to perform correction on the printing data PMN, to obtain corrected printing data PMN′. Specifically, because drive control data of a common printhead is binary, a printing instruction received by the printhead is actually 0 or 1. A correction process is that, when a value of the printing data PMN is 0, a value of the corrected printing data PMN′ that is obtained after the correction is performed is 1, so that the printhead can identify and perform the printing. More specifically, the first printing material is used for printing, provided that the printhead identifies the corrected printing data PMN′. In a specific embodiment, for example, the printing data PMN is data based on the CMY color mode, the printing data PMN is transferred to the printhead by using four transmission channels. Three of the four transmission channels are configured to transmit data of the CMY color mode, and the remaining one transmission channel is configured to transmit the corrected printing data PMN′. A controller of the printhead separately identifies the printing data PMN and the corrected printing data PMN′, and further starts to use the second printing material or the first printing material to perform the printing. In a variation, dedicated data channels are separately disposed for the printing data PMN and the corrected printing data PMN′, and a controller data channel matches and interconnects each of PMN and PMN′. That is, there are four controllers that respectively correspond to three types of data of the CMY color mode and the corrected printing data PMN′. In this case, a process of identifying the printing data PMN and the corrected printing data PMN′ may be omitted, and the printing may be started when the printing data PMN and the corrected printing data PMN′ are generated.

In a variation, an association is further established between the corrected printing data PMN′ and a printing startup instruction, and when the corrected printing data PMN′ is identified, the first printing material is used for printing. A person skilled in the art understands that, further, a printing process including multiple printheads is used in this variation. In a specific correction process, not only the correction is performed on the value of the printing data PMN, but also an association is established between the corrected printing data PMN′ that is obtained each time the correction is performed and printing startup instructions of different printheads, to implement multi-printhead printing, and increase the printing speed.

Further, step S105 is performed. Results of the layer-by-layer printing in step S102 are superposed to form a 3D object. A person skilled in the art understands that this step is a forming step. Step S102 to step S104 relate to printing of a specific layer, but the target object has M layers in total. The M layers are all printed layer by layer by performing step S102 to step S104, and the M layers are superposed to form the 3D object. More specifically, the superposition in this step is not a step that is finally performed, but is a step that is performed with step S102 to step S104. That is, one layer is superposed after a printing result of the layer is completed by performing step S102 to step S104. Such superposition process is an accumulative process. Further, each result of the layer-by-layer printing that is formed by performing step S102 to step S104 is first cured and then superposed, to further improve the dimensional stability of the 3D object, and avoid penetration between the results of the layer-by-layer printing. In a variation, after being leveled, each result of the layer-by-layer printing is first cured and then superposed. According to the variation, the dimensional stability of the 3D object can be further improved, and the formed 3D object can have a more aesthetic appearance.

In a first embodiment of the present invention, the layered image data DM is obtained by performing lamination and unit-division processing on the entire target object. A person skilled in the art understands that, in this embodiment, the lamination and unit-division processing is performed on the entire target object. This can most likely implement perfect reproduction of the target object, and continuity in both a shape and a color of the target object can be better implemented. Specially, when the target object has a perspective effect, the printing is further performed according to this embodiment. A person skilled in the art understands that, a specific process of the lamination and unit-division is divided into two parts. A first part is to perform lamination on the entire target object, to obtain multiple sections and spatial coordinates of the sections. A second part is to perform unit-division processing on each section, determine the number of pixel points of each unit, and determine a color value of each pixel point according to a color of the unit.

Further, for the lamination processing in the first part, a lamination processing algorithm based on grouping and sorting and calculation of an intersection of opposite sides is implemented. The basic idea thereof is that, entire grouping and sorting is performed according to geometrical continuity of an STL model, a lamination relationship matrix is established, then opposite sides of a triangular patch in the relationship matrix are separately tracked to calculate an intersection, and finally section profile data is generated. Specifically, the first part is implemented in the following manner: a first step is to perform grouping and sorting. A person skilled in the art understands that, compared with the size of the triangular patch in the STL model, a lamination thickness of rapid prototyping processing is usually very small, and a triangular patch usually intersects with multiple adjacent laminated planes. Because of the geometrical continuity of the STL model, triangular patches intersecting with a laminated plane are also arranged sequentially. Therefore, except a few of laminated panels, a set of the triangular patches intersecting with adjacent laminated panels is successive. That is, a set of patches intersecting with two adjacent laminated panels is basically unchanged. A second step is to calculate the intersection of the opposite sides, to obtain the section profile data. A specific process is usually that, a series of sides and a group of slice planes that are parallel to each other are known, intersection points between each side and the group of planes are calculated, and all intersection points on a same layer are sequentially arranged according to a connection relationship of the intersection points, forming a closed section profile of each layer. During actual operation, the lamination steps are usually performed by using lamination software. The specific number of layers needs to be determined according to a spatial resolution of a final 3D object. A higher spatial resolution indicates a larger number of layers, and a lower spatial resolution indicates a smaller number of layers.

Further, for the unit-division processing of the second part, the number of pixel points of each unit is first determined. Specifically, the number needs to be determined according to a resolution required by each section. A higher resolution indicates a larger number of pixel points, and a lower resolution indicates a smaller number of pixel points. More specifically, a particular section is divided into multiple units, and each unit is a rectangular pixel point array consisting of pixel points. For example, three primary colors CMY (that is, cyan, magenta, and yellow) are used as an example, the particular section is first divided into planes of the three primary colors, unit-division is performed on a plane of each primary color, and each unit that is formed after the division is a rectangular pixel point array consisting of pixel points of a particular primary color. For another example, four primary colors CMYK (that is, cyan, magenta, yellow, and black) are used as an example, the particular section is first divided into planes of the four primary colors, unit-division is performed on a plane of each primary color, and each unit that is formed after the division is a rectangular pixel point array consisting of pixel points of a particular primary color. Secondly, a color value of each pixel point is determined. This is described in the foregoing specific implementation, and details are not described herein again.

In a second embodiment of the present invention, the printing data PMN in step S102 is obtained by performing half-tone conversion on the layered image data DM. A person skilled in the art understands that, in this embodiment, the half-tone conversion is performed on the color values of the NM pixel points that are defined based on the section on which the NM pixel points are located. Specifically, in fact, a process of the half-tone conversion is to convert a value of a particular pixel point into corresponding printing data. More specifically, a person skilled in the art understands that, in the 3D printing, when the half-tone conversion is performed, not only a two-dimensional color resolution of the particular section needs to be considered, but also a spatial color resolution of the entire target object needs to be considered. A reason is that the 3D object is finally formed after multiple sections are superposed, and a color resolution of each section may superpose each other and finally affect the spatial color resolution of the 3D object. Specifically, when the half-tone conversion is performed on each section, color resolution standards of the sections may be consistent, or may be inconsistent. This depends on the spatial color complexity of the 3D object. For example, a color resolution of a section in a central part of the 3D object may be relatively low, and a color resolution of a section that is close to a surface part of the 3D object may be relatively high. For another example, a color resolution of a section that is distant from a transparent part of the 3D object may be relatively low, and a color resolution of a section that is close to the transparent part of the 3D object may be relatively high.

In an embodiment, an example in which the layered image data DM is 8-bit data is used for description. A person skilled in the art understands that, when the layered image data DM is 8-bit data, values of XMN corresponding to DM are between 0 and 255. As shown in FIG. 2, the printing data PMN is obtained in the following manner:

Step S201 is performed. The value set XMN is divided by 255 to obtain a corresponding fitted value set XMN′. The fitted value set XMN′ is between 0 and 1. Specifically, a person skilled in the art understands that, the value set XMN consists the values of the NM pixel points, including x11, x12, x21, x22, . . . , and xMN, and the fitted value set XMN′ is formed corresponding to the value set XMN. That is, the fitted value set XMN′ consists of fitted values of the NM pixel points, that is, the fitted value set XMN′ consists of x11′, x12′, x21′, x22′, . . . , and xMN′.

Further, step S202 is performed. A threshold S is set. If xMN′ is less than S, converted pMN is 0; or if xMN′ is greater than S, converted pMN is 1, and the printing data PMN that is correspondingly formed consists of a series of pMN. Further, the threshold S is a value between 0.4 and 0.65. Specifically, the value of S is 0.4, 0.5, 0.55, 0.6, or 0.65.

In a third embodiment of the present invention, as shown in FIG. 3, the printing data PMN is obtained in the following manner:

Step S301 is performed. The value set XMN is divided by 255 to obtain a corresponding fitted value set XMN′. The fitted value set XMN′ sequentially consists of xM1′, xM2′, xM3′, . . . , and xMN′ according to an order of neighboring relationships of N pixel points of an Mth-layer image. Specifically, a person skilled in the art understands that the neighboring relationships are further neighboring relationships of N pixel points that are on a same section layer. For example, a fitted value set X1N′ of a first-layer image consists of x11′, x12′, x13′, . . . , and x1N′, a fitted value set X2N′ of a second-layer image consists of x21′, x22′, x23′, . . . , and x2N′, and so on.

Further, step S302 is performed. A threshold S is compared with xM1′. If xM1′ is less than S, pM1 is 0; or if xM1′ is greater than S, pM1 is 1. A person skilled in the art understands that step S302 is a basic step of this embodiment, and may also be referred to as a start step, and the following steps are performed based on step S302. During actual application, a position of the fitted value xM1′ corresponding to step S302 may be randomly generated, or may be specified, and a position of a pixel point corresponding to the fitted value xM1′ is used as a center, provided that the fitted value xM1′ is generated. Processing is performed on a neighboring pixel point that is not processed. That is, processing is sequentially performed on fitted values of corresponding pixel points according to the order. Specifically, once a position of a pixel point corresponding to the fitted value xM1′ is generated, positions of pixel points corresponding to subsequent fitted values are not randomly generated, and are generated according to a sequence and based on the neighboring relationship. Finally, a pixel point sequence is formed, and a fitted valve sequence corresponding to the pixel point sequence is xM2′, xM3′, . . . , and xMN′. For example, after the position of the pixel point corresponding to xM1′ is generated, positions of pixel points are sequentially generated rightward and downward.

Further, step S303 is performed. A difference EMN is calculated by subtracting pMN from xMN′, and a sum of xM(N+1)′ and EMN is calculated to obtain xM(N+1)″. A person skilled in the art understands that step S303 is actually an error processing process. This is a difference between this embodiment and the second embodiment. An objective of step S303 is that, the half-tone conversion process is error-prone, and by means of the error processing, the color fidelity of the converted printing data PMN may be high. Specifically, this step is performed based on step S302, and the difference EMN spreads according to a sequence order of the pixel points in step S302. That is, an error of an adjacent point is accepted starting from xM2′. For example, the difference EM1 is obtained by subtracting pM1 from xM1′, and a method for accepting an error of xM1′ by xM2′ is adding xM2′ to EM1 to obtain xM2″, and using xM2″ as a processed object. Similarly, xM3′, xM4′, . . . , and xMN′ are sequentially caused to accept errors of neighboring fitted values.

Further, step S304 is performed. xM(N+1)″ is compared with the threshold S. If xM(N+1)″ is less than S, pM(N+1) is 0; or if xM(N+1)″ is greater than S, pM(N+1) is 1. Specifically, as show in FIG. 3, step S303 and step S304 are circularly performed. The number of cycles is determined according to a value of N until xMN′ is all converted into pMN. That is, x11′ is converted into p11, x12′ is converted into p12, x21′ is converted into p21, x22′ is converted into p22, . . . , until xMN′ is converted into pMN. The printing data PMN that is correspondingly formed consists of a series of pMN.

In a fourth embodiment of the present invention, as shown in FIG. 4, the printing data PMN is obtained in the following manner:

Step S401 is performed. The value set XMN is divided by 255 to obtain a corresponding fitted value set XMN′. The fitted value set XMN′ consists of a series of xMN′.

Further, step S402 is performed. The fitted value set XMN′ is sorted according to N pixel points of an Mth-layer image, and the fitted value set XMN′ sequentially consists of xM1′, xM2′, xM3′, xM4′, . . . , and xMN′. Specifically, a person skilled in the art understands that the sorting manner is further based on N pixel points that are on a same section layer. For example, after the sorting, a fitted value set X1N′ of a first-layer image consists of x11′, x12′, x13′, . . . , and x1N′, a fitted value set X2N′ of a second-layer image consists of x21′, x22′, x23′, . . . , and x2N′, and so on. Specifically, on each section layer, the sorting may be performed based on a pixel point set of a single unit, or the sorting may be performed based on a pixel point set of multiple units. A specific sorting manner is changeable, may be random sorting, or may be sequential sorting.

Further, step S403 is performed. g thresholds are set and sorted, to form a threshold set SMg′. A person skilled in the art understands that the threshold set SMg′ is set to make comparison with the fitted value set XMN′. The threshold set SMg′ includes the g thresholds. That is, the threshold set SMg′ sequentially consists of s1′, sM2′, sM3′, sM4′, . . . , and sMg′. Specifically, g represents the number of the thresholds included in the threshold set SMg′. A value of g may be equal to the number N of the pixel points in step S402, or may be less than the number N of the pixel points in step S402.

Step S404 is performed. Based on a sequence of the threshold set SMg′, N values xMN′ of the Mth-layer image in the fitted value set XMN′ are correspondingly compared with the g thresholds in the threshold set SMg′ one by one. If xMN′ is less than sMg′, pMN is 0; or if xMN′ is greater than sMg′, pMN is 1, and the printing data PMN that is correspondingly formed consists of a series of pMN. Specifically, this step is also an error processing process. This is a difference between this embodiment and the second embodiment. That is, the fitted value set XMN′ is not compared only with a fixed threshold S, but is compared with the threshold set SMg′. In this way, the diversity of the comparison can be implemented, and errors are manually reduced by setting the threshold set SMg′. An objective of step S404 is the same as that of the third embodiment, and details are not described herein again. According to the description of step S403, when g is equal to N, according to a sequence of the threshold set SMg′, that is, according to an order of sM1′, sM2′, sM3′, sM4′, . . . , and sMg′, comparison is sequentially made with corresponding xM1′, xM2′, xM3′, . . . , and xMN′ one by one, to obtain a value of pMN. When g is less than N, the comparison is still made one by one according to the sequence of the threshold set SMg′. A difference is that, because g is less than N, the N values of the Mth-layer image in the fitted value set XMN′ cannot be all compared in one comparison process. In this case, cyclic comparison needs to be performed based on the sequence of the threshold set SMg′, that is, multiple comparison processes are included. Specifically, first, a first comparison process is completed, and sM1′, sM2′, sM3′, sM4′, . . . , and sMg′ are compared with xM1′, xM2′, xM3′, xM4′, . . . , and xMg′; if 2*g<N, a second comparison process is started, and sM1′, sM2′, sM3′, sM4′, . . . , and sMg′ are compared with xM(g+1)′, xM(g+2)′, xM(g+3)′, xM(g+4)′, . . . , and xMN; and if 2*g>N, a third comparison process is continued, and sM1′, sM2′, sM3′, sM4′, . . . , and xMg′ are compared with xM(g+1)′, xM(g+2)′, xM(g+3)′, xM(g+4)′ . . . xM(2*g); then a fourth comparison process is performed, to compare the remaining fitted value xMN′ one by one in a similar manner. A person skilled in the art understands that N may be greater than several times of g. Therefore, during actual application, a fifth comparison process, a sixth comparison process, and the like may also be included. The comparison manner is similar, and details are not described herein again. More specifically, further, a value range of the threshold set SMg′ is any value between 0.4 and 0.65.

Further, in the third embodiment and the fourth embodiment, the layered image data DM may alternatively be 16-bit data, and a value of XN corresponding to DM is between 0 and 65535. A specific implementation process is similar to that of the third embodiment and the fourth embodiment, and details are not described herein again.

According to another aspect of the present invention, a system for printing a 3D object is provided, including a data processor, a process controller, and a printhead. The data processor converts layered image data DM of a target object to obtain printing data PMN, and performs correction on data being 0 in the printing data PMN to obtain corrected printing data PMN′. DM consists of a value set XMN of NM pixel points. A person skilled in the art understands that the data processor usually includes a central processing unit, a main memory, and an input/output interface. The data processor is configured to process data, and is a part that executes an instruction according to a step specified in a procedure. A common operating process is that, the central processing unit reads the layered image data DM from the main memory by using the input interface, converts the layered image data DM according to a particular procedure, obtains the printing data PMN and the corrected printing data PMN′ after the conversion, and outputs the printing data PMN and the corrected printing data PMN′ by using the output interface.

Further, the process controller usually includes a central processing unit, a main memory, and an input/output interface. The processor controller is configured to convert particular data into an executable operating instruction, stores the executable operating instruction in the main memory, reads the operating instruction during running to form a control signal, and allocates the signal to each part for execution. A common operating process is that, the central processing unit receives the printing data PMN and the corrected printing data PMN′, by using the input interface, generates a particular printing instruction according to the printing data PMN and the corrected printing data PMN′, and stores the instruction in the main memory. A printing instruction for performing printing by using a first printing material is generated according to the corrected printing data PMN′, and a printing instruction for performing printing by using a second printing material is generated according to the printing data PMN. Finally, the central processing unit converts the printing instruction into a control signal, and outputs the signal to the printhead by using the output interface. The printhead performs layer-by-layer printing according to the control signal, to form a 3D object by means of superposition. In another variation, the process controller directly outputs the generated printing instruction to the printhead by using the output interface, instead of converting the printing instruction into a control signal. In this case, the printhead matches the received printing instruction with a built-in instruction of the printhead. If the matching succeeds, the printhead performs printing; or if the matching fails, the printhead does not perform printing.

In a preferable embodiment, the system further includes a leveling apparatus and a curing apparatus. The leveling apparatus is configured to perform leveling on each result of the layer-by-layer printing. Further, the leveling apparatus includes a transmission apparatus and a function apparatus. The function apparatus may be a leveling roller or a squeegee, and the transmission apparatus may be a mechanical arm. The curing apparatus is configured to perform curing on each result of the layer-by-layer printing. Because the first printing material is different from the second printing material, the curing apparatus may be a thermal curing apparatus, or may be a light-curing apparatus.

The foregoing describes the specific embodiments of the present invention. It should be understood that, the present invention is not limited to the foregoing particular implementations, and a person skilled in the art may make various variations or modifications within the scope of the claims. This cannot affect essential content of the present invention.

Read more
PatSnap Solutions

Great research starts with great data.

Use the most comprehensive innovation intelligence platform to maximise ROI on research.

Learn More

Patent Valuation

46.08/100 Score

Market Attractiveness

It shows from an IP point of view how many competitors are active and innovations are made in the different technical fields of the company. On a company level, the market attractiveness is often also an indicator of how diversified a company is. Here we look into the commercial relevance of the market.

32.0/100 Score

Market Coverage

It shows the sizes of the market that is covered with the IP and in how many countries the IP guarantees protection. It reflects a market size that is potentially addressable with the invented technology/formulation with a legal protection which also includes a freedom to operate. Here we look into the size of the impacted market.

39.18/100 Score

Technology Quality

It shows the degree of innovation that can be derived from a company’s IP. Here we look into ease of detection, ability to design around and significance of the patented feature to the product/service.

43.21/100 Score

Assignee Score

It takes the R&D behavior of the company itself into account that results in IP. During the invention phase, larger companies are considered to assign a higher R&D budget on a certain technology field, these companies have a better influence on their market, on what is marketable and what might lead to a standard.

25.74/100 Score

Legal Score

It shows the legal strength of IP in terms of its degree of protecting effect. Here we look into claim scope, claim breadth, claim quality, stability and priority.

PatSnap Solutions

PatSnap solutions are used by R&D teams, legal and IP professionals, those in business intelligence and strategic planning roles and by research staff at academic institutions globally.

PatSnap Solutions
Search & Analyze
The widest range of IP search tools makes getting the right answers—and asking the right questions—easier than ever. One click analysis extracts meaningful information on competitors and technology trends from IP data.
Business Intelligence
Gain powerful insights into future technology changes, market shifts and competitor strategies.
Workflow
Manage IP-related processes across multiple teams and departments with integrated collaboration and workflow tools.
Contact Sales