EP4186027A1 - Procédé d'inspection de qualité et agencement d'inspection de qualité - Google Patents
Procédé d'inspection de qualité et agencement d'inspection de qualitéInfo
- Publication number
- EP4186027A1 EP4186027A1 EP21769368.8A EP21769368A EP4186027A1 EP 4186027 A1 EP4186027 A1 EP 4186027A1 EP 21769368 A EP21769368 A EP 21769368A EP 4186027 A1 EP4186027 A1 EP 4186027A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- dimension
- image recording
- sub
- areas
- way
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000007689 inspection Methods 0.000 title claims abstract description 15
- 238000010146 3D printing Methods 0.000 claims abstract description 7
- 239000000843 powder Substances 0.000 claims description 34
- 238000004220 aggregation Methods 0.000 claims description 32
- 230000002776 aggregation Effects 0.000 claims description 32
- 239000011248 coating agent Substances 0.000 claims description 21
- 238000000576 coating method Methods 0.000 claims description 21
- 230000007547 defect Effects 0.000 claims description 20
- 239000000654 additive Substances 0.000 claims description 7
- 230000000996 additive effect Effects 0.000 claims description 7
- 238000004519 manufacturing process Methods 0.000 claims description 7
- 238000010894 electron beam technology Methods 0.000 claims description 6
- 239000000853 adhesive Substances 0.000 claims description 5
- 230000001070 adhesive effect Effects 0.000 claims description 5
- 239000010410 layer Substances 0.000 description 31
- 238000001514 detection method Methods 0.000 description 14
- 238000011161 development Methods 0.000 description 9
- 230000018109 developmental process Effects 0.000 description 9
- 230000008901 benefit Effects 0.000 description 8
- 238000007639 printing Methods 0.000 description 6
- 238000013459 approach Methods 0.000 description 5
- 238000009877 rendering Methods 0.000 description 5
- 238000011835 investigation Methods 0.000 description 4
- 239000002184 metal Substances 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000001454 recorded image Methods 0.000 description 4
- 230000015572 biosynthetic process Effects 0.000 description 2
- 238000005755 formation reaction Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000004026 adhesive bonding Methods 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 239000011230 binding agent Substances 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000002844 melting Methods 0.000 description 1
- 230000008018 melting Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000012768 molten material Substances 0.000 description 1
- 238000013021 overheating Methods 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 239000002356 single layer Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
- G06T7/001—Industrial image inspection using an image reference approach
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B22—CASTING; POWDER METALLURGY
- B22F—WORKING METALLIC POWDER; MANUFACTURE OF ARTICLES FROM METALLIC POWDER; MAKING METALLIC POWDER; APPARATUS OR DEVICES SPECIALLY ADAPTED FOR METALLIC POWDER
- B22F10/00—Additive manufacturing of workpieces or articles from metallic powder
- B22F10/30—Process control
- B22F10/38—Process control to achieve specific product aspects, e.g. surface smoothness, density, porosity or hollow structures
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B22—CASTING; POWDER METALLURGY
- B22F—WORKING METALLIC POWDER; MANUFACTURE OF ARTICLES FROM METALLIC POWDER; MAKING METALLIC POWDER; APPARATUS OR DEVICES SPECIALLY ADAPTED FOR METALLIC POWDER
- B22F10/00—Additive manufacturing of workpieces or articles from metallic powder
- B22F10/80—Data acquisition or data processing
- B22F10/85—Data acquisition or data processing for controlling or regulating additive manufacturing processes
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B22—CASTING; POWDER METALLURGY
- B22F—WORKING METALLIC POWDER; MANUFACTURE OF ARTICLES FROM METALLIC POWDER; MAKING METALLIC POWDER; APPARATUS OR DEVICES SPECIALLY ADAPTED FOR METALLIC POWDER
- B22F12/00—Apparatus or devices specially adapted for additive manufacturing; Auxiliary means for additive manufacturing; Combinations of additive manufacturing apparatus or devices with other processing apparatus or devices
- B22F12/90—Means for process control, e.g. cameras or sensors
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B29—WORKING OF PLASTICS; WORKING OF SUBSTANCES IN A PLASTIC STATE IN GENERAL
- B29C—SHAPING OR JOINING OF PLASTICS; SHAPING OF MATERIAL IN A PLASTIC STATE, NOT OTHERWISE PROVIDED FOR; AFTER-TREATMENT OF THE SHAPED PRODUCTS, e.g. REPAIRING
- B29C64/00—Additive manufacturing, i.e. manufacturing of three-dimensional [3D] objects by additive deposition, additive agglomeration or additive layering, e.g. by 3D printing, stereolithography or selective laser sintering
- B29C64/30—Auxiliary operations or equipment
- B29C64/386—Data acquisition or data processing for additive manufacturing
- B29C64/393—Data acquisition or data processing for additive manufacturing for controlling or regulating additive manufacturing processes
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B33—ADDITIVE MANUFACTURING TECHNOLOGY
- B33Y—ADDITIVE MANUFACTURING, i.e. MANUFACTURING OF THREE-DIMENSIONAL [3-D] OBJECTS BY ADDITIVE DEPOSITION, ADDITIVE AGGLOMERATION OR ADDITIVE LAYERING, e.g. BY 3-D PRINTING, STEREOLITHOGRAPHY OR SELECTIVE LASER SINTERING
- B33Y40/00—Auxiliary operations or equipment, e.g. for material handling
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B33—ADDITIVE MANUFACTURING TECHNOLOGY
- B33Y—ADDITIVE MANUFACTURING, i.e. MANUFACTURING OF THREE-DIMENSIONAL [3-D] OBJECTS BY ADDITIVE DEPOSITION, ADDITIVE AGGLOMERATION OR ADDITIVE LAYERING, e.g. BY 3-D PRINTING, STEREOLITHOGRAPHY OR SELECTIVE LASER SINTERING
- B33Y50/00—Data acquisition or data processing for additive manufacturing
- B33Y50/02—Data acquisition or data processing for additive manufacturing for controlling or regulating additive manufacturing processes
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B22—CASTING; POWDER METALLURGY
- B22F—WORKING METALLIC POWDER; MANUFACTURE OF ARTICLES FROM METALLIC POWDER; MAKING METALLIC POWDER; APPARATUS OR DEVICES SPECIALLY ADAPTED FOR METALLIC POWDER
- B22F10/00—Additive manufacturing of workpieces or articles from metallic powder
- B22F10/10—Formation of a green body
- B22F10/14—Formation of a green body by jetting of binder onto a bed of metal powder
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B22—CASTING; POWDER METALLURGY
- B22F—WORKING METALLIC POWDER; MANUFACTURE OF ARTICLES FROM METALLIC POWDER; MAKING METALLIC POWDER; APPARATUS OR DEVICES SPECIALLY ADAPTED FOR METALLIC POWDER
- B22F10/00—Additive manufacturing of workpieces or articles from metallic powder
- B22F10/20—Direct sintering or melting
- B22F10/28—Powder bed fusion, e.g. selective laser melting [SLM] or electron beam melting [EBM]
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B29—WORKING OF PLASTICS; WORKING OF SUBSTANCES IN A PLASTIC STATE IN GENERAL
- B29C—SHAPING OR JOINING OF PLASTICS; SHAPING OF MATERIAL IN A PLASTIC STATE, NOT OTHERWISE PROVIDED FOR; AFTER-TREATMENT OF THE SHAPED PRODUCTS, e.g. REPAIRING
- B29C64/00—Additive manufacturing, i.e. manufacturing of three-dimensional [3D] objects by additive deposition, additive agglomeration or additive layering, e.g. by 3D printing, stereolithography or selective laser sintering
- B29C64/10—Processes of additive manufacturing
- B29C64/141—Processes of additive manufacturing using only solid materials
- B29C64/153—Processes of additive manufacturing using only solid materials using layers of powder being selectively joined, e.g. by selective laser sintering or melting
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B33—ADDITIVE MANUFACTURING TECHNOLOGY
- B33Y—ADDITIVE MANUFACTURING, i.e. MANUFACTURING OF THREE-DIMENSIONAL [3-D] OBJECTS BY ADDITIVE DEPOSITION, ADDITIVE AGGLOMERATION OR ADDITIVE LAYERING, e.g. BY 3-D PRINTING, STEREOLITHOGRAPHY OR SELECTIVE LASER SINTERING
- B33Y10/00—Processes of additive manufacturing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20016—Hierarchical, coarse-to-fine, multiscale or multiresolution image processing; Pyramid transform
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20021—Dividing image into blocks, subimages or windows
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30144—Printing quality
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P10/00—Technologies related to metal processing
- Y02P10/25—Process efficiency
Definitions
- the invention relates to a quality inspection method according to claim 1 and an arrangement for a quality inspection according to claim 10 .
- additive manufacturing machines so-called “3D printers” usually use cameras to monitor the layer-by-layer printing process. This occurs in particular in the field of metal printing, for example when using the so-called “Selective Laser Melting” or the so-called sand printing, such as the so-called “binder jetting”.
- the quality of the respective new coating is usually monitored by a powder bed camera, which is used to detect missing or inhomogeneous powder scattering.
- EP 3 459 715 A1 which relates to a prediction of post-coating problems in the laser powder bed melt.
- WO 2015/020939 A1 tries to improve this, which uses a comparison between a camera image and a rendering for detection.
- the rendering is generated from a CAD model by a so-called "slicer”.
- the object on which the invention is based is therefore to provide a solution that overcomes the disadvantages of the prior art.
- This task is based on the quality inspection method according to the preamble ff of claim 1 and by the Arrangement for quality inspection according to claim 10 each solved by the characterizing features.
- a new coating with a powder takes place for each layer according to the virtual layer to be printed, b.
- the new coating is fixed, in particular with a laser, adhesive or electron beam, c.
- At least one first digital image recording of the fixed layer is generated, d.
- at least one second digital image recording of the new coating is generated, e.
- At least one virtual division of the first image recording and at least one virtual division of the second image recording takes place in such a way that the first image recording and the second image recording each because it is subdivided into sub-areas in such a way that all sub-areas cover the entire image at least once per image, and the first division is carried out in such a way that a sub-area of the first image is a sub-area of the second image in position and dimensions, as well as the sum of the parts areas of the first image recording corresponds to the sum of the partial areas of the second image recording, f.
- an examination is carried out in such a way that they are subjected to a correlation and aggregation, supported in particular by machine learning.
- the correlation and aggregation is repeated at least for each pair of sub-areas of the first image recording and the second image recording defined by the same position and the same dimension, with the position of each further pair of sub-areas successively in the x and/or y direction is shifted according to a Cartesian coordinate system, h . an error procedure is initiated if a defect is signaled at least on the current layer on the basis of at least one correlation and/or aggregation that has been carried out.
- the dimension which is generally defined as the extent of bodies, is defined in the context of this revelation in such a way that the size of the partial area is designated, which is defined, for example, by a first number of pixels in x and a second number is spanned by pixels in the y-direction according to a Cartesian coordinate system.
- the method according to the invention thus differs from a pure pixel-by-pixel comparison of images, since the examination is based on areas formed by the partial area.
- differences in brightness, for example, which falsify the gray values of the materials recorded do not or only to a much lesser extent false test results.
- This is further reinforced by the approach according to the invention of pairs of surfaces for receiving the layer coated with powder and for receiving the layer fixed, in particular by laser exposure, electron beam irradiation or gluing, because according to inventive considerations there are relationships between these layers ments that can be worked out based on correlation with the investigation according to the invention and so u. a. Virtually eliminate falsifications caused by external influences.
- the image before the coating is compared with the image after the coating, so that missing powder in this before/after comparison can basically be recognized by the fact that both images show locally similar contents. Because, according to the invention, not only the correlation is carried out, but also an aggregation, this is strengthened and also more accurate. This is achieved because, according to the invention, adjacent partial areas of the recordings of the fixed and newly coated layer are also fed to the correlation. According to considerations according to the invention, adjacent sub-areas also have a relationship to one another, which is worked out by the correlation in connection with the aggregation in such a way that reliable indications of defects, such as missing powder, can be provided.
- these indications ie the detection of errors, are thus further improved in that, according to the invention, not only the correlation but also an aggregation takes place as a result.
- D. H the results of the correlations per area are subjected to a collective analysis in a suitable manner, so that the named relationships of neighboring partial areas are worked out as error information.
- the arrangement according to the invention for the quality inspection for three-dimensional printing such as according to the so-called "additive manufacturing” method, in which a three-dimensional print object is generated in such a way that the print object is virtually divided into layers and successive successive layers of the print model by the printers are printed in such a way that a. for this purpose a new coating with a powder takes place for each layer according to the virtual layer to be printed, b. the new coating is fixed, in particular with a laser, adhesive or electron beam, has means for carrying out the method according to the invention.
- the virtual division of the first image recording and the virtual division of the second image recording preferably takes place in such a way that first sub-areas are formed which have a first dimension and second sub-areas are formed which have a second dimension, the second dimension is greater than the first , a .
- the correlation takes place in such a way that a correlation coefficient is formed for each pair of first partial surfaces located within a pair of second partial surfaces, with the position of each additional pair of partial surfaces being successively in the x and/or y direction is shifted according to a Cartesian coordinate system, b .
- the aggregation takes place in such a way that for each second partial area the sum of the correlation coefficients is formed over all first pairs of first partial areas positioned within the second partial area, c. the error method, at least based on a comparison of the sum with a first threshold value initiated when the threshold is exceeded.
- the first dimension is defined as part of this revelation as the size of the first partial area, which is given, for example, by the area that spans a number of pixels in the x and y direction
- the second dimension is defined as part of this revelation Size of the second sub-area is defined, which is given, for example, by the number of first sub-areas in the x-direction and the number of first sub-areas in the y-direction, ie a larger number of pixels in the x- and y-directions.
- the first sub-areas thus define cells, so to speak, which form a subset of second sub-areas, which can be viewed as blocks, since the second sub-areas are larger in terms of dimension, ie the area under consideration.
- the pairs of cells and the pairs of blocks are formed for the correlation and aggregation, with the cells being shifted step by step within the blocks because the blocks are larger until the entire area to be examined, which, for example, corresponds at most to the dimension of the corresponds to the block, ie the area of the block, has contributed to the investigation at least once.
- the shifting of the cells can, for example, also take place until the entire area of the recording for examination has been covered at least once by a cell, i.e. the entire area has contributed to the correlation and subsequently the division of the entire area of the recording into blocks containing a certain number of cells can be carried out in order to carry out the aggregation.
- the new pairs formed by the shift are also included in the investigation.
- This division into blocks and cells also means that, according to inventive considerations, these have relationships to one another which, through suitable correlation and aggregation, enable the accurate detection of errors and support.
- a coefficient is formed for each pair of the first sub-area and this is then added up to the respective coefficient of the coefficient formed by the shifted first sub-area (cells) within a block.
- This summation forms the aggregation and leads to a total value, i.e. a collected consideration, which, if it exceeds a threshold value, represents an indication of an error.
- the threshold value is selected in such a way that this signal is accurate.
- this value can be set or optimized beforehand and/or during further operation, for example by empirical values, parameters of the environment and/or, in particular by machine learning-supported, optimization procedure.
- the correlation coefficient is preferably formed as a so-called Pearson correlation coefficient. This is particularly suitable for working out relationships and for aggregation.
- the coefficient is normalized before the aggregation in such a way that when it falls below a second threshold value, in particular a value less than 0. 9, the coefficient is set to the value 0, is a threshold value, specified for example in the sense of an optimal default value, and thus also supports the implementation of the method or method according to the invention. indicates a possible implementation of the threshold .
- the invention is further developed in such a way that a ratio of the first dimension to the second dimension is defined in such a way that at least two first partial areas overlap, then, according to the inventive consideration, the relationships between the pairs of first partial areas used for the investigation are further reinforced since the neighboring i.e. new pairs of cells and/or blocks created by shifting the position are then no longer disj unkt. It is also advantageous here, among other things, that the overlapping parts, ie surfaces, are subjected to the examination several times.
- Local interference such as light reflections, which falsify the respective recording, in particular the gray value level of the pixels, may disappear in the next cell and can reveal a concealed error or avoid a false error message, but at least through the collected overview give a hint, for example , da individual values have a stronger or weaker effect on correlation and/or aggregation .
- the development representing the implementation is given if the first dimension and second dimension are defined by the number of pixels in the x and y direction according to a Cartesian coordinate system, the ratio of the first dimension to the second dimension is defined in such a way that that at least two first sub-areas for at least one pixel row have an overlap, in particular the first dimension is defined by 3 x 3 pixels and the second dimension is defined by 5 sub-areas x 5 sub-areas, whereby this is ultimately also a division according to pixels
- Subdivision according to pixels makes particular sense when considering the resolution of the digital camera that creates the image recordings, since this is also related to pixels and optimal values, such as cell, block and/or threshold sizes, can thus be determined as simply as possible.
- the quality inspection f becomes more granular.
- the definition of "adjacent" in the context of this revelation also includes cells that are not disjointed ) and two rows of pixels in the y-direction, and is therefore an advantageous choice, for example in terms of an optimal default value for an implementation, since a relatively large area overlaps with, among other things, the above-mentioned advantage of compensating for locally occurring interference.
- the ratio of the first dimension defined by the number of pixels and the second dimension is preferably optimized in such a way that it particularly increases the resolution of the first and second camera that creates images and/or other parameters of the three-dimensional print.
- the aforementioned optimization, taking into account the camera resolution, is hereby specifically applied to the dimension ratio and thus contributes to the above-mentioned advantages in this regard.
- the examination is carried out on the basis of partial areas determined by at least a first pair of a first recorded image and a second recorded image and a second pair of a first recorded image and a second recorded image.
- FIGURE 1 is an image of a layer with multiple missing powder defects after recoating according to the prior art, marked by the yellow box.
- FIG. 2 images of layers as they appear after exposure according to the prior art
- FIG. 3 shows a flow chart of an exemplary embodiment of the method according to the invention
- FIG. 4 shows schematic examples of the correlation and aggregation according to the invention
- FIG. 5 shows a schematic representation of examples of error detection based on an exemplary embodiment of the invention.
- FIGURE 1 is an image of the coating of a layer with powder as it occurs in 3D printing according to the prior art ("post-coating image") of a layer with multiple defects of missing powder (“Powderf ehlde ekten”), which are marked by a frame , see how it shows according to the prior art. It can be seen that defects near the upper left corner are difficult to see due to different degrees of reflectance, resulting in very low contrast in this area.
- FIGURE 1 for the recording of a new coating with powder with several missing powder defects shows that due to the different degrees of reflection that can result depending on the path and lighting of the image acquisition for the recording, the bare metal , i.e. the places where the Powder is missing, recorded with very different gray values, which makes it considerably more difficult to identify the defects in the low-contrast regions, such as in the upper half of the framed area in FIG.
- FIGURE 2 These variations are exemplified in FIGURE 2 .
- FIGURE 2 shows images that result after exposure.
- after exposure means that a laser has selectively melted the powder from the post-coating.
- the gray value variation is mainly caused by differences in light reflection and generally does not convey any information about the process quality. However, it makes image analysis more difficult as the melted areas can be represented by very different gray levels, sometimes indistinguishable even from powder gray levels.
- FIG. 3 shows a flowchart of an embodiment of the method according to the invention, which meets these challenges.
- a first recording of the currently fixed print layer in the 3D printer takes place in a first step S 1 .
- the layer of powder is effected by exposure to a laser.
- the invention is not restricted to this. Rather, the application of the invention alternative methods for fixing can be done, such as the method in which the powder is bonded by means of an adhesive or is fixed by means of an electron beam.
- a first and a second digital recording are thus available. One before and one after coating with the powder. Each of these two recordings is then virtually subdivided into blocks in a third step S3.
- the division can be carried out in such a way that the total area of each recording, in particular in equally dimensioned frames enclosing, for example, consecutive partial areas of the area of the recording - the blocks - is separated in such a way that they form a partial area of the recording, for example for an assignment, define .
- the division for the two recordings takes place in the same way, so that a partial surface of the first recording produced in this way has the same position, shape and dimensions as the partial surface of the second recording produced in this way.
- a fourth step S 4 an identically dimensioned and positioned first cell is then formed in each block at a starting position for each recording, ie for the first and also the second recording. Also virtual, as characterized in the way previously described.
- a second cell for each recording is then formed in a fifth step S5; So they overlap with each with the starting line .
- the second cells correspond to the first cell in terms of dimensions and shape, but are slightly offset in terms of position in the x-direction.
- a further pair of cells is also formed for each recording starting from the first cell formed in the fourth step S 4 , ie a third cell for each recording.
- These are also identical to the first cell in terms of dimension and shape, but their position is offset in relation to the first cell in the y-direction. So they also overlap the first cell.
- the degree of overlap of the two dislocations can be the same, but is not limited to this.
- a seventh step S7 a check is now made as to whether the end of the block has already been reached after the second cell pair has been formed by the first cell pair and the second cell pairs and third cell pairs; the entire block was thus covered virtually by a first pair of cells and at least a second pair of cells and at least a third pair of cells. If this is not the case after this first offset, the fifth step S5, the sixth step S6 and the seventh step S7 are repeated, so that a further second and third cell pair is formed.
- the number of shifted cells in the y-direction is equal to that in the x-direction.
- the example and the invention are not limited to this. Different offset steps can take place in one direction and in the other. For the example shown, this means that, as a possible variant, either the extent of the offset in the x-direction can be selected differently than the extent in the y-direction. However, a further variant also consists in the steps being able to take place in a different order and/or in separate loops.
- the cell pairs can be correlated in an eighth step S 8 according to the invention.
- this correlation is carried out in a subroutine and all of the first cell pair and the second and third cell pairs are transferred to this subroutine.
- the invention is not restricted to this implementation variant; rather, as noted above, the structure of the implementation can be completely different. For example, individual pairs could be transferred to the subroutine, or subroutines are not used, but alternative implementations are used to correlate the pairs.
- a ninth step S 9 an aggregation takes place over all of the cell pairs. Details on aggregation and correlation are given based on the schematic representation of an exemplary embodiment of these processes in FIGURE 4 .
- the block is then classified. This is shown as one step for the sake of simplicity, but could also be implemented as a subroutine that is called with the results of the aggregation or, as explained above, completely different implementations for implementing the idea according to the invention.
- the contribution of the exemplary embodiment to solving the problem on which the invention is based consists in a suitable correlation between at least two images, in which, in order to detect missing powder, one image is taken after exposure and one after recoating.
- the inventive contribution to the solution is based on the inventive idea that if the powder is missing, the local patterns in the post-coating image correlate highly with the exposure image in the regions where the powder is missing.
- the correlation according to the exemplary embodiment of the invention requires that sliding windows, the previously described cells, are moved synchronously over the images to be compared.
- cells are rather small, for example 3 x 3 pixels or 5 x 5 pixels in size, in order to also detect small defects. to be able to grasp.
- the optimal cell size will be determined as a function of the spatial resolution of the digital camera and will then generally remain constant for a specific type of printer.
- the Pearson correlation coefficient of the respective sample ie the respective cell pair, is then calculated for each cell position as follows where ca, cb denote the same cell in images a and b, pk denote the pixel values of the cell, and m denotes the average pixel value of the cell.
- the numerator in Eq. (1) measures the common variation of the pixel values of the cell pk around the respective cell mean value m in the cell of both images. If the cell shows the same variation value in both images, i.e. a pixel is larger (or smaller) than the respective cell mean value in both images, then the product term in the sum of the numerator is positive. It is negative for an opposite variation, for example in image a the pixel value is greater than the cell mean while the same pixel in b is smaller than the mean. The sum runs over all pixels of the cell.
- the denominator normalizes the sum by the product of the standard deviations of the cells.
- the result of this step is a correlation coefficient value r 6 [-1, 1] for each cell. Values close to +1 indicate a high correlation (similarity) of the respective cell in both images.
- the aggregation schematically indicated by the larger rectangle in the two images in the figure now requires the application of a threshold value to all correlation coefficients.
- a threshold value for example, all values smaller than 0.9 can be set to zero.
- the resulting values are then aggregated across neighboring cells. This is achieved by defining floating cells that overlap by one pixel, which together result in a block of 5 by 5 cells indicated as a larger rectangle, whose values are added up.
- a high total per block at the end of this process indicates a defect in the missing powder in that block.
- further criteria can be included in the aggregation/correlation, for example a requirement that the local variance per cell exceeds at least a constant minimum value. In this way, erroneous correlations can be avoided, which can arise, for example, due to similar JPEG compression artifacts in both images, which are irrelevant with regard to powder errors, but can influence local correlations.
- FIG. 5 shows examples of successful detection of missing powder when using the exemplary embodiments according to the invention, which therefore successfully show an error.
- the use of small cells for the correlation enables the detection of smaller defects.
- the small cells can also lead to high correlations in benign cases, ie in the absence of defects. This can occur, for example, when molten material of an exposed area partially protrudes from the powder, so that it is partially visible in the recoating image.
- the solution according to the invention thus offers, through the combination described, a means for detecting powder defects which is robust and at the same time highly sensitive to small defects, which is one of the most critical.
- the solution according to the invention also produces fewer false alarms.
- the invention also exhibits an invariance with respect to local brightness fluctuations, for example due to inhomogeneous lighting.
- the exemplary embodiments of the invention are also distinguished by easily interpretable and easily tunable threshold values, since the correlation coefficients are limited by 1.0, in particular when the threshold value is set to 1.
- the invention is advantageously very well suited for implementing fast architectures from common CNN frameworks, such as what is known as “tensor flow”.
Landscapes
- Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Materials Engineering (AREA)
- Manufacturing & Machinery (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Automation & Control Theory (AREA)
- Mechanical Engineering (AREA)
- Optics & Photonics (AREA)
- Analytical Chemistry (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP20192906.4A EP3961557A1 (fr) | 2020-08-26 | 2020-08-26 | Procédé d'inspection de la qualité et agencement pour une inspection de qualité |
PCT/EP2021/073312 WO2022043283A1 (fr) | 2020-08-26 | 2021-08-24 | Procédé d'inspection de qualité et agencement d'inspection de qualité |
Publications (1)
Publication Number | Publication Date |
---|---|
EP4186027A1 true EP4186027A1 (fr) | 2023-05-31 |
Family
ID=72243038
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
EP20192906.4A Pending EP3961557A1 (fr) | 2020-08-26 | 2020-08-26 | Procédé d'inspection de la qualité et agencement pour une inspection de qualité |
EP21769368.8A Pending EP4186027A1 (fr) | 2020-08-26 | 2021-08-24 | Procédé d'inspection de qualité et agencement d'inspection de qualité |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
EP20192906.4A Pending EP3961557A1 (fr) | 2020-08-26 | 2020-08-26 | Procédé d'inspection de la qualité et agencement pour une inspection de qualité |
Country Status (4)
Country | Link |
---|---|
US (1) | US20230360194A1 (fr) |
EP (2) | EP3961557A1 (fr) |
CN (1) | CN116600919A (fr) |
WO (1) | WO2022043283A1 (fr) |
Family Cites Families (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9390494B2 (en) * | 2012-12-13 | 2016-07-12 | Kla-Tencor Corporation | Delta die intensity map measurement |
US9855698B2 (en) | 2013-08-07 | 2018-01-02 | Massachusetts Institute Of Technology | Automatic process control of additive manufacturing device |
JP6295561B2 (ja) * | 2013-09-17 | 2018-03-20 | 株式会社リコー | 画像検査結果判断装置、画像検査システム及び画像検査結果の判断方法 |
ZA201505683B (en) * | 2014-08-15 | 2017-11-29 | Central Univ Of Technology Free State | Additive manufacturing system and method |
WO2017087451A1 (fr) * | 2015-11-16 | 2017-05-26 | Materialise N.V. | Détection d'erreurs dans des processus de fabrication additifs |
EP3459715A1 (fr) | 2017-09-26 | 2019-03-27 | Siemens Aktiengesellschaft | Procédé et appareil pour prédire l'apparition et le type de défauts dans un processus de fabrication additive |
DE102019102484A1 (de) * | 2019-01-31 | 2020-08-06 | Carl Zeiss Smt Gmbh | Verarbeitung von dreidimensionalen Bilddatensätzen |
-
2020
- 2020-08-26 EP EP20192906.4A patent/EP3961557A1/fr active Pending
-
2021
- 2021-08-24 WO PCT/EP2021/073312 patent/WO2022043283A1/fr active Application Filing
- 2021-08-24 US US18/022,365 patent/US20230360194A1/en active Pending
- 2021-08-24 CN CN202180073265.2A patent/CN116600919A/zh active Pending
- 2021-08-24 EP EP21769368.8A patent/EP4186027A1/fr active Pending
Also Published As
Publication number | Publication date |
---|---|
WO2022043283A1 (fr) | 2022-03-03 |
EP3961557A1 (fr) | 2022-03-02 |
US20230360194A1 (en) | 2023-11-09 |
CN116600919A (zh) | 2023-08-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
DE69515646T2 (de) | Verfahren und Vorrichtung zum Erfassen von Fehlern in Textilbahnen und dergleichen | |
DE102006000946B4 (de) | Verfahren und System zur Inspektion einer periodischen Struktur | |
DE4410603C1 (de) | Verfahren zur Erkennung von Fehlern bei der Inspektion von strukturierten Oberflächen | |
DE60100594T2 (de) | Verfahren und system zur aufspürung von fehlern auf einer gedruckten leiterplatte | |
DE3336471A1 (de) | Verfahren und anordnung zur inspektion von bildern | |
EP1132732A2 (fr) | Procédé pour évaluer des criques structurales sur la surface d'une plaquette semiconductrice | |
EP2888110B1 (fr) | Procédé d'inspection d'au moins un exemplaire d'un produit d'impression | |
EP3970113B1 (fr) | Détermination du niveau d'usure d'un outil | |
DE102019208149A1 (de) | Nozzle Health mit neuronalem Netz | |
EP2605212A2 (fr) | Procédé et dispositif de vérification optique d'objets à vérifier lors de la fabrication et/ou de l'emballage de cigarettes | |
DE102020126554A1 (de) | Mikroskopiesystem und verfahren zum überprüfen von eingabedaten | |
DE102016005173A1 (de) | Verfahren zur Prüfung von Zigaretten oder Zigarettenpackungen | |
EP4123506A1 (fr) | Procédé et dispositif d'analyse d'un produit, procédé d'apprentissage, système, programme informatique et support d'enregistrement lisible par ordinateur | |
DE102007025304A1 (de) | Verfahren zur Verbesserung der Reproduzierbarkeit einer Koordinaten-Messmaschine und deren Genauigkeit | |
EP2787485A1 (fr) | Procédé et dispositif de détection automatique d'erreurs dans des corps flexibles | |
DE19951146A1 (de) | Verfahren zum Reduzieren des Rauschens in einem durch Abbildung erhaltenen Signal | |
WO2022037899A1 (fr) | Procédé et appareil de fabrication additive d'une pièce à travailler | |
WO2022043283A1 (fr) | Procédé d'inspection de qualité et agencement d'inspection de qualité | |
EP3518180B1 (fr) | Procédé de création d'une structure de données de piles d'images | |
DE69026488T2 (de) | Verfahren zur Qualitätsbeurteilung von Druckerzeugnissen | |
DE102008052248B4 (de) | Verfahren und Vorrichtung für die Überprüfung von Linsenstrukturen | |
DE60033200T2 (de) | Verfahren und Vorrichtung zur Zuordnung von Wafern | |
DE102008033171A1 (de) | Verfahren und Vorrichtung zur Inline-Qualitätssicherung an Druckmaschinen | |
EP3762870B1 (fr) | Élément de télémétrie | |
EP1347286A1 (fr) | Procédé et appareil d'inspection d'objets |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: UNKNOWN |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE |
|
PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE |
|
17P | Request for examination filed |
Effective date: 20230221 |
|
AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR |
|
DAV | Request for validation of the european patent (deleted) | ||
DAX | Request for extension of the european patent (deleted) | ||
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: EXAMINATION IS IN PROGRESS |
|
17Q | First examination report despatched |
Effective date: 20240209 |