Light homogenization method for multi-source large-scale surface exposure 3D printing

Abstract

A light homogenization method for multi-source large-scale surface exposure 3D printing, comprising the following steps: projecting pure-color images of a first color and a second color having identical attributes capturing an image of an overlapping portion and calculating height and width information of the overlapping portion; splitting a pre-processed slice and respectively recording width and height information of two slices resulting from the splitting and generating two grayscale images having identical attributes thereto; counting power values of identical positions of slices in different grayscale values, performing a further calculation to obtain a projection mapping function, using the projection mapping function as a basis for performing optimization on grayscale interpolation of the generated images; and fusing the processed grayscale images and the originally split two slices to obtain a surface exposure 3D printing slice having a uniform brightness in final shaping.

Claims

1. An energy homogenization method for multiple-projector large-scale mask projection 3D printing comprising the steps of: Step 100: using at least two projectors as a light-source for mask projection, and locating the projectors to be adjacent to each other so that there is an overlapped area between the projection areas of the projectors, wherein the two projectors project two images with the same property but with different pure colors, a first color and a second color, and the overlapped area between the projection areas of the projectors has a third color; then, using a camera to capture an image of the projection areas and the overlapped area, and analyzing the image using a computer, wherein the height and width of the overlapped area are denoted as H.sub.0 and W.sub.0 respectively; Step 200: based on the information of the height H.sub.0 and the width W.sub.0 of the overlapped area, segmenting a preprocessed slice which is obtained from the captured image to create two slices which are denoted as P.sub.1 and P.sub.2 respectively; recording the width W.sub.1 and W.sub.2 and the H.sub.1 and H.sub.2 of the two slices P.sub.1 and P.sub.2; and then generating two corresponding gray leveled pictures P.sub.3 and P.sub.4 having the same property with the two slices; Step 300: measuring the output energy at the same position on a printing area with discrete gray levels; by analyzing the measured data, obtaining a mapping function T[r(x, y)] by using curve fitting; and based on the mapping function T[r(x,y)], optimizing the generated pictures P.sub.3 and P.sub.4 using gray level interpolation; and Step 400: fusing the pictures P.sub.3 and P.sub.4 treated above with the two slices P.sub.1 and P.sub.2 to generate a series of 3D printed slices with energy homogenization.

2. The method of claim 1, wherein Step 100 comprises the substeps of: testing and adjusting the levelness of the projectors by using a leveler so that the projectors have same projection orientation; then fixing these projectors so that their relative position is invariable; activating the projectors to project images with different pure colors so that the overlap area between them has the third color which is a mixture of the two pure colors and can be identified by a computer; and using the camera to capture the image of the projection areas, including the areas with the two pure colors and that with the mixed color; obtaining details of the overlapped area based on the differences between image pixels; and recording the height and width of the overlapped area as H.sub.0 and W.sub.0 respectively.

3. The method of claim 1, wherein Step 200 comprises the substeps of: zooming the slices to the size of the printing area with the invariable aspect ratio, and recording the height and width of the whole slice as H and W respectively; segmenting the slice based on the height H.sub.0 and width W.sub.0 of the overlapped area, recording the segmented slices as P.sub.1 and P.sub.2, and recording their widths as W.sub.1 and W.sub.2 and their heights as H.sub.1 and H.sub.2; and generating two gray leveled images P.sub.3 and P.sub.4 with the same gray levels attributes with P.sub.1 and P.sub.2 based on the widths and heights of the two segmented slices P.sub.1 and P.sub.2.

4. The method of claim 1, wherein in Step 300: for the power value at the same position with different gray levels, the complete projection mapping function is obtained through Fourier series curve fitting:
T[r(x,y)]=a.sub.0+a.sub.1*cos(r(x,y)*w)+b.sub.1*sin(r(x,y)*w) wherein r(x, y) is the gray level at location (x, y), w is angular frequency, and a.sub.0 and a.sub.1 are constants.

5. The method of claim 4, wherein according to the relationship of the overlapped area and projection mapping function of gray leveled images P.sub.3, P.sub.4, the illumination energy in the printing area of the gray leveled images P.sub.3, P.sub.4 are determined based on the following energy homogenization formula: $\{\begin{array}{c}{r}^{}\left(x,y\right)=\mathrm{argmin}\left(\underset{\left(x,y\right)s_{\max }}{.Math.}{\left(f\left(T_1\left[r_1\left(x,y\right)\right]+T_2\left[r_2\left(x,y\right)\right]\right)-f\right)}^2-f\right)\\ f\left(T_1+T_2\right)=\{\begin{array}{c}{T_1\left[r_1\left(x,y\right)\right]}_{\left(x,y\right)S_1}\\ {T_2\left[r_2\left(x,y\right)\right]}_{\left(x,y\right)S_2}\\ {T_1\left[r_1\left(x,y\right)\right]}_{\left(x,y\right)S_3}+{T_2\left[r_2\left(x,y\right)\right]}_{\left(x,y\right)S_3}\end{array}\end{array}$ wherein section S1 is defined as an area that belongs to gray image P.sub.3 without overlapped with P.sub.4, section S2 is defined as an area that belongs to gray images P.sub.4 without overlapped with P.sub.3, section S.sub.3 is defined as the overlapped area between gray images P.sub.3 and P.sub.4, S.sub.max is the maximum exposure area, and f is the average energy in the whole exposure area.

6. The method of claim 5, wherein for the energy homogenization formula, illumination unevenness of the exposure areas in sections S.sub.1, S.sub.2 and S.sub.3 is reduced in the following way: 1) dividing sections S.sub.1, S.sub.2 and S.sub.3 into MN image sub-blocks respectively; for sections S.sub.1 and S.sub.2, using areas in the slice image as exposable areas; finding corresponding energy values from the obtained sub-areas of the image of the candidate area for exposure area; and selecting the minimum energy value as the optimal target energy value in the exposure area; 2) obtaining the illumination energy corresponding to the gray value of each pixel at the boundary of sections S.sub.1 and S.sub.3, and storing it in an array A; obtaining the illumination energy corresponding to the gray value of each pixel of the boundary of sections S.sub.2 and S.sub.3, and storing it in an array B; establishing two linear equations reflecting the change in the height or width in sections S.sub.3; and determining the energy value of each position in section S.sub.3 based on the combination of the energy values of the two linear equations at the same position.

7. The method of claim 6, wherein after the energy value of each position in the third section S.sub.3 is determined, linear interpolation is performed in sections S.sub.1 and S.sub.2 respectively so that two gray level images with smooth gray changes are obtained.

8. The method of claim 7, wherein in Step 400: the gray value of each pixel in the interpolated gray level image is sequentially scanned; the next pixel is skipped if the gray level value is zero; if the gray value is greater than zero, the gray value of the pixel is obtained, and then the gray value is assigned to the same pixel position of the original image slice; and finally, the pixels of the segmented slices P.sub.1 and P.sub.2 are distributed in gray levels to satisfy the pixel gray distribution of the gray images P.sub.3 and P.sub.4, respectively.

9. The method of claim 1, wherein by using a plurality of projectors of the same energy distribution as light sources of the mask projection 3D printer, Steps 100 to 400 and the corresponding sub-steps are performed for every two projectors adjacent in the height direction or/and in the width direction.

10. The method of claim 9, wherein for the energy values of the projected gray level images in the same position in different gray levels values, a complete projection mapping function is obtained through Fourier series curve fitting:
T[r(x,y)]=a.sub.0+a.sub.1*cos(r*w)+b.sub.1*sin(r*w) wherein r(x, y) is a gray image at location (x, y), w is the angular frequency, and a.sub.0 and a.sub.1 are constants.

11. The method of claim 10, wherein the illumination power in the exposable areas of each slice is determined according to the relationship of the intersection positions of the slices and the projection mapping function and based on the following light power formula: $\{\begin{array}{c}{r}^{}\left(x,y\right)=\mathrm{argmin}\left(\underset{\left(x,y\right)s_{\max }}{.Math.}{\left(f\left(\underset{i=1}{\overset{n}{.Math.}}{T}_n\left[r_n\left(x,y\right)\right]\right)-f\right)}^2-f\right)\\ f\left(\underset{i=1}{\overset{n}{.Math.}}{T}_n\left[r_n\left(x,y\right)\right]\right)=\{\begin{array}{c}{T_n\left[r_n\left(x,y\right)\right]}_{\left(x,y\right)S_n}\\ {\underset{i=1}{\overset{n}{.Math.}}{T}_n\left[r_n\left(x,y\right)\right])}_{\left(x,y\right)S_m}\end{array}\end{array}$ wherein f is the average energy at the locations in the printing area, S.sub.n is the sections without overlap, and S.sub.m is the overlapped section between the projectors.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 is a schematic diagram of 3D printing, which takes two projectors as example.

(2) FIG. 2 is a flow chart of the energy homogenization method for large scale mask projection 3D printing using multiple projectors of the present invention.

(3) FIG. 3 (a) shows the original slice, and (b) shows the upper and lower parts of the slices after segmentation.

(4) FIG. 4 is an illustration of labeling information in the projection image.

(5) FIG. 5 is an illustration of the projection mapping function curve fitting.

(6) FIG. 6 shows the energy distribution of a projector for the image gray level of 255.

(7) FIGS. 7 (a) and (b) respectively show the distribution of energy of the upper and lower slice images after using the method.

(8) FIG. 8 shows the distribution of energy in the whole slice region using this method.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

(9) Some embodiments of the disclosure will be described with reference to the drawings. It should be noted that, various embodiments and various features of the embodiments of the disclosure can be implemented in combination in the condition that no confliction is caused.

(10) The invention provides an energy homogenization method for large scale mask projection 3D printing using multiple projectors. By dealing with each slice, the interference is reduced and large-area exposure is achieved. The large area mentioned is determined by the number of specific projections, i.e., the width and height are considered to be at least 280 mm280 mm.

(11) The mask projection 3D printer mainly uses a digital light processing (DLP) projector as a light source, and the most important component in the projector is a digital micro-mirror device (DMD) to complete the visual digital information display technology. Specifically, DLP projection technology uses DMD chips as the primary key processing element to implement digital optical processing. However, if a larger surface is required and the high power density of the curing light is to be satisfied at the same time, the light intensity must be greatly increased. But DMD does not withstand the high light intensity. At the same time, the light intensity increases, the system cooling problem is serious. Therefore, the current 3D printing based on DLP technology is developing slowly in the processing format. Therefore, we have designed an algorithm for large-area mask projection printing in which multiple projectors as the light sources of mask projection 3D printer are used for exposure, which solves the problem of splicing between slices that are projected by each projector. FIG. 1 a schematic diagram of 3D printing, which takes two projectors for example. This schematic diagram uses two projectors 1 and 2 as examples. Wherein, the upper slice 1010 is projected by the upper projector, the lower slice 1020 is projected by the lower projector, the overlapping portion 1030 is located in the overlapped area, and the camera is 1040.

(12) Here, it should be pointed out that, for the convenience of description, in the example shown in the figure, two projectors 1, 2 are stacked one above the other to produce an upper slice, a lower slice, and an overlapped area between them. It is understood that the principle of the present invention is also applicable to the case where two projectors are overlapped on the left and right sides to generate a left half of the slice, a right half of the slice, and an overlapped area between them. It will also be understood that the principles of the present invention are also applicable to multiple projector combinations. For a large width and height, it may be necessary to put two or more projectors in the width and height directions together to implement the mosaic effect of the multi-source exposure surface. The projectors are preferably arranged in a matrix, and there are overlapped areas between the adjacent slices (left and right) (i.e., in the widthwise direction) and the adjacent slices (up and down) (i.e., in the height direction).

(13) The drawings are used to explain the basic principle of the present invention. It is described for the examples of two projectors placed side by side in the figures. It can be understood that the described features are also applicable to projectors arranged in other ways or other numbers of projectors.

(14) FIG. 2 is a flow chart of an energy homogenization method for large scale mask projection 3D printing using multiple projectors for the embodiment shown in FIG. 1.

(15) The invention provides an energy homogenization method for large scale mask projection 3D printing using multiple projectors, includes:

(16) Step 100: Ensure that two projectors of the same specification are placed side by side under the condition of just full contact. Two projectors output images with red and green respectively. The upper image is red and the down image is green. And the overlapped region is the yellow. Using a camera to capture the image of the projection area and the overlapped area could be obtained by analyzing the image using computer, the height and width of the overlapped areas are denoted as H.sub.0 and W.sub.0 respectively;

(17) Step 200: According to the information of the overlapped region, and segment the preprocessed slice. And the two slices can be denoted as P.sub.1 and P.sub.2. Meanwhile, width and height of the P.sub.1 and P.sub.2 are denoted as W.sub.1 and W.sub.2, H.sub.1 and H.sub.2 respectively. Then two gray leveled images P.sub.3 and P.sub.4 are generated with the same properties of P.sub.1 and P.sub.2.

(18) Step 300: Measuring the output energy at the same position with some discrete gray levels. By analyzing the statistic data, the mapping function T[r(x, y)] is acquired by using curve fitting. Based on the mapping function T[r(x, y)], the power value of gray leveled pictures at the same position with different gray levels is calculated. The generated pictures P.sub.3 and P.sub.4 are optimized based on the projection mapping functions.

(19) Step 400: When pictures P.sub.3 and P.sub.4 have been completed, fusing P.sub.3 (and P.sub.4) with P.sub.1 and (P.sub.2), so the slices with energy homogenization are generated.

(20) Among them, the processing Step 100 includes:

(21) Substep 110: The levelness of projector 1 and projector 2 is tested by using the leveler, so that the projectors have same projection orientation. The projector 1 and projector 2 are fixed so that their relative position is invariable.

(22) Substep 120: Let the two projectors 1 and 2 respectively project red and green pure color images. Because the two projectors are placed in a stack, the projection areas must overlap, allowing one projector to project a red image and one to project a green image. This ensures that the overlapping area is yellow and is easily recognized by the computer. Making these projectors project a different pure color image and try to make the overlapped area easily to identify.

(23) Substep 130: Using a camera to capture the image of the projection area including the areas with two pure color and the mixed color. Based on the differences between image pixels, the detail of the overlapped area are easily obtained. The height and width of the overlapped area are denoted as H.sub.0 and W.sub.0 respectively.

(24) Further, Step 200 includes:

(25) Substep 210: Zooming the slices to the size of the printing area with the invariable aspect ratio. In addition, the exposed parts of the slices should have same size of projectors' project image. And record the whole slice's height as H, and the width is W.

(26) Substep 220: According to the height H.sub.0 and width W.sub.0 of the overlapped area, the slices will be segmented using the following formula. And the segmented slices are denoted as P.sub.1 and P.sub.2. Their widths are W.sub.1 and W.sub.2, and heights are H.sub.1 and H.sub.2:
H.sub.1=H.sub.2=H/2+H.sub.0(1)
W=W.sub.0=W.sub.1=W.sub.2(2)

(27) Among them, FIG. 3(a) shows the original slice, and (b) shows the upper and lower parts of the slices after segmentation, and FIG. 4 is illustration of labeling information in the projection image.

(28) Substep 230: According to the width and height of P.sub.1 and P.sub.2 in the two sections after segmentation, two gray leveled images P.sub.3 and P.sub.4 are generated with the same gray levels attributes of P.sub.1 and P.sub.2.

(29) Further, Step 300 includes:

(30) Substep 310: Measuring the output energy at the same position with some discrete gray levels and find that the energy distribution of the projector is similar at different gray levels at the same position, after further fitting experiments, it was found that the power is non-linearly changing and the law conforms to the Fourier series fitting distribution. The confidence interval is 95%. A complete projection mapping function can be obtained by curve fitting:
T[r(x,y)]=a.sub.0+a.sub.1*cos(r*w)+b.sub.1*sin(r*w)(3)
where r(x, y) is the corresponding brightness of the picture at different positions. FIG. 5 is illustration of the projection mapping function curve fitting. In the above equation, the r represents the gray levels, the w represents the angular frequency, and both the a.sub.0 and the a.sub.1 represent constants.

(31) Substep 320: According to the relationship between the intersecting positions of the pictures P.sub.3 and P.sub.4 and the projection mapping function, the expression of the problem shown in the following formula can be obtained. To solve the problem of uneven distribution of illumination energy in the exposable areas (including overlapping areas) of the pictures P.sub.3 and P.sub.4. The portion that belongs to the picture P.sub.3 and does not intersect with the picture P.sub.4 is the first portion S.sub.1, the portion that belongs to the picture P.sub.4 and does not intersect with the picture P.sub.3 is the second portion S.sub.2, and the portion that overlaps with the pictures P.sub.3 and P.sub.4 is the overlapping portion S.sub.3, and the S.sub.max is the maximum exposure area, as follows:

(32) $\begin{array}{cc}\{\begin{array}{c}{r}^{}\left(x,y\right)=\mathrm{argmin}\left(\underset{\left(x,y\right)s_{\max }}{.Math.}{\left(f\left(T_1\left[r_1\left(x,y\right)\right]+T_2\left[r_2\left(x,y\right)\right]\right)-f\right)}^2-f\right)\\ f\left(T_1+T_2\right)=\{\begin{array}{c}{T_1\left[r_1\left(x,y\right)\right]}_{\left(x,y\right)S_1}\\ {T_2\left[r_2\left(x,y\right)\right]}_{\left(x,y\right)S_2}\\ {T_1\left[r_1\left(x,y\right)\right]}_{\left(x,y\right)S_3}+{T_2\left[r_2\left(x,y\right)\right]}_{\left(x,y\right)S_3}\end{array}\end{array}& \left(4\right)\end{array}$
where f is the average energy in the whole exposure area, and the number of light sources involved is greater than the case where two units are n, the expression of the problem can also be written as follows:

(33) $\begin{array}{cc}\{\begin{array}{c}{r}^{}\left(x,y\right)=\mathrm{argmin}\left(\underset{\left(x,y\right)s_{\max }}{.Math.}{\left(f\left(\underset{i=1}{\overset{n}{.Math.}}{T}_n\left[r_n\left(x,y\right)\right]\right)-f\right)}^2-f\right)\\ f\left(\underset{i=1}{\overset{n}{.Math.}}{T}_n\left[r_n\left(x,y\right)\right]\right)=\{\begin{array}{c}{T_n\left[r_n\left(x,y\right)\right]}_{\left(x,y\right)S_n}\\ {\underset{i=1}{\overset{n}{.Math.}}{T}_n\left[r_n\left(x,y\right)\right])}_{\left(x,y\right)S_m}\end{array}\end{array}& \left(5\right)\end{array}$
where S.sub.n denotes the portion of each projector that does not overlap with other projectors, and S.sub.m denotes the overlap between the projectors, as the number of projectors is multiplied, the number of areas where the exposure of the projector overlaps is also increasing, and each overlapping part may be generated by overlapping the projections of adjacent projectors in the width direction and/or height direction.

(34) Substep 330: In order to solve the problem of unequal illumination in the exposable area of each section S.sub.1, S.sub.2, and S.sub.3, in formula 4, we use the following method to meet the requirements of the above formula to solve the problem: 1) sections S.sub.1, S.sub.2, and S.sub.3 are divided into MN image sub-blocks respectively. For sections S.sub.1 and S.sub.2, the areas in the slice image are used as exposable areas. Finding the corresponding energy from the obtained sub-areas of the image of the candidate area for exposure area. and the minimum energy is selected as the optimal target energy in the exposure area; 2) Get the lower boundary of the first part S.sub.1 (that is, the boundary that falls in the second part S.sub.2). The illumination energy corresponding to the gray value of each pixel is stored in the array A. Obtain the upper boundary of the second portion S.sub.2 (that is, the boundary that falls in the first portion S.sub.1). The illumination energy corresponding to the gray value of each pixel is stored in the array B. When W.sub.1=W.sub.2, two linear equations with the slopes of K.sub.1 and K.sub.2 with the height change in the overlap area S.sub.3 as independent variables are respectively established, where K.sub.1 and K.sub.2 are respectively represented as:
K.sub.1=a[W.sub.1]/(H.sub.1H.sub.0)(6)
K.sub.2=b[W.sub.2]/(H.sub.0H.sub.2)(7)

(35) According to the superposition of the power values of these two linear equations at the same position, it is the power value of each position in the overlapping portion S.sub.3 area.

(36) It is understood that for the case where two projectors are placed side by side in the width direction, a linear equation with two slopes can be separately established with the variation of the width in the area of the overlapping portion S.sub.3 as an independent variable.

(37) For the case where the number of light sources is larger than two projectors, the overlapping portion S.sub.3 may be a superposition of projection images of three or even four projectors. In this regard, linear equations with three or four slopes can be established for the variation of height and width in the area of overlap S.sub.3 as independent variables, respectively, then, based on the superposition of the energy values of the three or four linear equations at the same position, the energy value at each position of the overlapping portion S.sub.3 is obtained.

(38) Substep 340: Linear interpolation the two gray leveled images that have been generated, then the two gray leveled pictures with smooth gray changes can be obtained. FIG. 6 shows the projection energy irradiance distribution of the projector when the gray levels of the image is 255. FIGS. 7(a) and (b) respectively show the distribution of energy of the upper and lower slice images after using the method.

(39) Step 400 includes:

(40) Substep 410: Sequentially scans the gray value of each pixel in the interpolated gray level image and skips the next pixel if the gray level value is zero. If the gray value is greater than zero, obtain the gray value of the pixel, then the gray value is assigned to the same pixel position of the original image slice. Finally the pixels of the segmented slices P.sub.1 and P.sub.2 are distributed in gray levels to satisfy the pixel gray distribution of the gray images P.sub.3 and P.sub.4, respectively. FIG. 8 shows the distribution of energy in the whole slice region using this method.

(41) In the example described above, two projectors respectively project red and green pictures. However, it can be understood that the present invention is applicable in a more general sense to an image projected by an adjacent projector having a first color (pure color) and a second color (pure color) different from the first color. The first color and the second color have a significant color difference (for example, two of the three primary colors are used), but the attributes are the same, and the color of the overlapping area between the two is a third color that is different from the first and second colors.

(42) Here, the picture attributes of the first color and the second color are the same, meaning that their size information and resolution information are the same.

(43) In addition, examples of two projectors placed side by side above and below are described above, and the present invention is not limited to the specific examples and details described, instead, after modifying the previously described details, it can be applied to situations where projectors or other numbers of projectors are arranged in other ways.

(44) An embodiment of the invention provides an energy homogenization method for large scale mask projection 3D printing using multiple projectors has the following advantages: 1) Improve the scale of exposure area; 2) Portabilitywhen the projection mapping function of different light sources is acquired, the method can be easily transplanted; 3) Printabilitythis method of energy homogenization of multi-source large-scale mask projection 3D printing can be applied to most models that do not need to be optimized. It has strong applicability and high success rate of one-time printing. Therefore, the present invention has a certain application value and significance.