WO2006136600A1 - Trame de similigravure a modulation de frequence multiniveau et procede permettant de produire cette trame - Google Patents

Trame de similigravure a modulation de frequence multiniveau et procede permettant de produire cette trame Download PDF

Info

Publication number
WO2006136600A1
WO2006136600A1 PCT/EP2006/063461 EP2006063461W WO2006136600A1 WO 2006136600 A1 WO2006136600 A1 WO 2006136600A1 EP 2006063461 W EP2006063461 W EP 2006063461W WO 2006136600 A1 WO2006136600 A1 WO 2006136600A1
Authority
WO
WIPO (PCT)
Prior art keywords
threshold
image
matrix
values
halftoning
Prior art date
Application number
PCT/EP2006/063461
Other languages
English (en)
Inventor
Joris Van Garsse
Original Assignee
Agfa Graphics Nv
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Priority claimed from EP05105659A external-priority patent/EP1737211A1/fr
Application filed by Agfa Graphics Nv filed Critical Agfa Graphics Nv
Priority to JP2008517509A priority Critical patent/JP2008544667A/ja
Publication of WO2006136600A1 publication Critical patent/WO2006136600A1/fr

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
    • H04N1/40—Picture signal circuits
    • H04N1/40087—Multi-toning, i.e. converting a continuous-tone signal for reproduction with more than two discrete brightnesses or optical densities, e.g. dots of grey and black inks on white paper

Definitions

  • the present invention relates to a digital image processing method. More specifically the invention relates to a method for digital halftoning of continuous tone images for printing purposes. Even more specifically, the invention relates to frequency modulation halftoning techniques that use threshold masks. Even more specifically, the invention relates to frequency modulation halftoning techniques that support printing devices that can print two or more density levels per pixel.
  • InkJet printing works by jetting ink droplets through a nozzle onto a substrate. By applying mechanical or thermal energy to ink in a small chamber, a pressure wave is created that propels a miniscule ink droplet at high speed through the nozzle towards a substrate.
  • the pressure wave is controlled by shaping the length and the profile of the electrical waveform that is applied to the thermal or mechanical transducer in the ink chamber.
  • tone resolution of such a printing system is usually too limited to render high quality graphics and continuous tone images.
  • Digital halftoning techniques are generally subdivided into two classes: amplitude modulation halftoning (also called clustered dot halftoning) and frequency modulation halftoning (also called dispersed dot halftoning) .
  • amplitude modulation halftoning also called clustered dot halftoning
  • frequency modulation halftoning also called dispersed dot halftoning
  • Clustered dot halftoning relies on a geometrical arrangement of halftone dots that is characterized by a set of fixed frequencies and angles. The tone values are modulated by changing the size of the halftone dots on the grid.
  • Frequency modulation halftoning relies on changing the average distance between essentially fixed sized halftone dots to modulate the tone values .
  • Hybrid forms of digital halftoning also exist wherein both the size or density and the average distance of halftone dots are modulated.
  • binary halftoning is commonly used to refer to halftoning techniques that are adapted for devices that can render only two tone levels per pixel corresponding to printing ink or no ink. These devices are referred to as binary printing devices.
  • multilevel halftoning is commonly used to refer to halftoning techniques that are adapted for devices that can render more two tone levels per pixel corresponding to printing with various densities of ink or with various sizes of ink spots. These devices are referred to as multilevel printing devices.
  • frequency modulation has been the preferred digital halftoning method in inkjet printing.
  • the class of frequency modulation halftoning techniques itself can be further subdivided into a first subclass that uses an algorithmic approach - called error diffusion - to convert a continuous tone image into a frequency modulated halftone, and a second subclass that uses pre-calculated threshold masks for this purpose.
  • a desirable characteristic of a frequency modulation halftoning technique is that it produces dot distributions that have "blue noise" spectrum, i.e. of which the power spectrum has less energy at lower frequencies than at higher frequencies. Error diffusion algorithms are known to have this desirable property and therefore enable to produce halftone dot distributions of excellent image quality.
  • a disadvantage of the error diffusion algorithm is that it requires relatively many computations because of the complexity of the algorithm. For high speed image processing this translates into the requirement of developing dedicated computing hardware, which is costly and takes time to develop.
  • Another problem is that, in the case of multilevel halftoning, the error diffusion algorithm offers few options to control the trade off between modulating the number of halftone dots and the density levels of said halftone dots.
  • US Pat. No. 5,214,517 by Sullivan et all describes the use of a set of e.g. 256 pre-calculated bitmaps, each one corresponding to a tone level ranging from 0 to 255 in a continuous tone image.
  • Each bitmap represents a distribution of halftone dots to render the corresponding tone level, and is calculated - using an iterative optimization process - to produce a power spectrum with minimal visual noise.
  • the bitmaps are typically square having a size of e.g. 256 by 256, and are replicated (i.e.
  • bitmaps horizontally and vertically across the two-dimensional addressable printer space like tiles on a floor.
  • a major disadvantage of this approach is that it requires a lot of memory. If the size of the bitmaps is 256 by 256 pixels, and 256 bitmaps are needed for as many tone levels, than 2 Mbytes of memory have to be provided.
  • Another disadvantage of this method is that the bitmaps can not be used in standard Raster Image Processors (RIP' s) that are used for the rendering of PostScript® of PDF® documents, because these devices perform halftoning using threshold masks.
  • RIP' s Raster Image Processors
  • Mitsa and Parker disclose a technique to build a square threshold mask that is replicated horizontally and vertically to yield a threshold value at every pixel position of a page for the purpose of frequency modulation halftoning.
  • Such a square threshold mask is compatible with standard Raster Image Processors, and requires only one byte per element. By comparing every pixel in the image with a corresponding threshold in the threshold mask, it is determined if at the corresponding location a halftone dot is printed or not.
  • the way the square threshold mask is calculated in the previous method is by accumulating successive levels of stochastic dot patterns on top of each other. Every higher level uses the dot distribution of the lower level plus additional dots. To determine which dots are to be added, a blue noise filter is applied on the dot distribution using the Fourier transform. By comparing the original dot distribution with the filtered dot distribution, it is determined at which locations dots should be added to optimally preserve the blue noise characteristics.
  • the threshold mask is obtained this way produces halftone dot distributions that have blue noise characteristics.
  • Multilevel techniques using threshold masks have been described in the art, for example in the document US 5,903,713 by Daels . These techniques, however, are mostly described in the context of periodic halftoning. The specific problems and solutions related to stochastic multilevel halftoning techniques, however, are not addressed in the documents relating to the state of the art.
  • One of these problems is a need to control the balance between changing the number of halftone dots and changing the density of halftone dots when reproducing various densities in an image. From the above analysis of the state of the art, it follows that a need exists for an alternative and improved method to calculate a threshold matrix for the purpose of frequency modulation digital halftoning.
  • Figure 1 shows a print head having nozzles.
  • Figure 2 shows the formation of ink droplets having a variable volume according to an embodiment of the current invention.
  • Figure 3 shows a printing system according to an embodiment of the current invention.
  • Figure 4 shows a print head comprising multiple columns of nozzles.
  • Figure 5 shows a print head assembly comprising multiple print heads .
  • Figure 6 shows a data processing system according to an embodiment of the current invention.
  • Figure 7 shows a printer controller according to an embodiment of the current invention.
  • Figure 8 shows an arrangement that uses a threshold array to convert a continuous tone image into a halftoned image according to an embodiment of the current invention.
  • Figure 9 explains the principle of converting a continuous tone image into a binary halftoned image according to an embodiment of the current invention.
  • Figure 10 explains the principle of converting a continuous tone image into a multilevel halftoned image according to a first embodiment of the current invention.
  • Figure 11 explains the principle of converting a continuous tone image into a multilevel halftoned image according to a second embodiment of the current invention to calculate a single threshold mask.
  • Figure 12A and 12B show a flow chart to explain a method according to an embodiment of the current invention.
  • Figure 16 shows a preferred embodiment of a single values function
  • Figure 17 shows an example of a parent matrix according to an embodiment of the current invention.
  • Figure 18 shows an example of a set of children matrices derived form a parent matrix in Figura 17 using one of the methods according to the current invention.
  • Figure 19 shows a look up table to transform values of an order number table according to an embodiment of the current invention.
  • Figure 20 shows a flow chart to explain a method according to an embodiment of the current invention to calculate multiple threshold masks .
  • Figure 21 shows a flow chart to explain a method according to an embodiment of the current invention to calculate multiple threshold masks .
  • Figure 22 explains different steps to calculate multiple threshold masks according to an embodiment of the current invention.
  • Figure 23 explains different steps to calculate multiple threshold masks according to an embodiment of the current invention.
  • the method according to the arrant invention is directed towards the use in inkjet printers and specifically drop-on-demand inkjet printers.
  • printing as used in the invention refers to the process of creating a structured pattern of ink markings on a substrate .
  • the ink could be a conventional pigmented or dyed ink or colorant, but it could also be wax, a water repellent substance, an adhesive or a plastic.
  • ink is not a pure compound, but a complex mixture comprising several components such as dyes, pigments, surfactants, binders, fillers, solvents, water and dispersants, each component serving a specific function.
  • the ink could also be a material of which, the viscosity or phasa changes with temperature, s ⁇ ch as wax.
  • the substrate could be paper, but it could also be textile, a synthetic foil or a metal plate or any other printable substrate as known in the art. Examples includes the use of inkjet printing (drop-on-demand and continuous) to create printing masters for offset printing, to print on packages in cardboard or plastic, to print decorative images.
  • the unprinted substrate can be available either on roll or as a pile of separate, usually rectangular sheets.
  • the substrate is characterized by a width called substrateWidth. When the substrate is on roll, this width corresponds with the width of the roll. When the substrate is available as a rectangular sheet, this width corresponds with one of the two dimensions of said rectangular sheet, for example the shortest dimension of said sheet.
  • the inkjet nozzles 130 can be operated in parallel and produce droplets with a fixed or a variable volume.
  • the volume of a droplet is modulated by means of a variable number of micro-droplets.
  • Figure 2 shows a case that the number of micro-droplets can vary between 0 and 7. Because the micro-droplets are so tiny, the air friction subjects them to substantial deceleration, and as a result, the micro-droplets coagulate into one single droplet having a variable volume by the time it lands on the printable substrate.
  • the volume of these droplets is referred to by the number of micro-droplets of which said droplet consists. More specifically in the example shown in Figure 2, the volume of a droplet having a variable size is referred to by an integer number ranging from 0 to 7 (represented by binary numbers in Figure 2) .
  • the tone resolution of a printing device shall refer to the number of density values that can be printed by a given device at a pixel position with a single ink.
  • printing an image of a document using a printer 300 is achieved by- moving a substrate 340 having an ink-receiving layer 341 relative to a page-wide print head assembly 350 by means of a substrate transport mechanism and selectively jetting ink droplets on said substrate in response to said image of said document.
  • the substrate 340 rests on a substrate support 310.
  • Figure 3 shows not one but multiple printing heads. In one embodiment four print heads (360, 361, 362, 363) are used that print with cyan, magenta, yellow and black ink. These heads are mounted on a print head support 350. Not shown on Figure 3 are the ink reservoirs and the tubes that are necessary for operating the printing device.
  • the inks spots on the substrate that result from droplets jetted by different heads are allowed to physically mix before they are cured. This technique of jetting subsequent droplets without intermediate curing is referred to as wet-on-wet printing.
  • Figure 3 also shows a curing station 370 for curing the printed inks after they have been printed on the substrate.
  • the curing station is a UV source that radiates energy having a spectrum that causes curing of the ink.
  • the orientation that corresponds with the movement of substrate relative to the print head is called the fast scan orientation.
  • the direction 330 in which the substrate moves relative to the print head shall be called the fast scan direction.
  • the orientation perpendicular the fast scan orientation is called the slow scan orientation.
  • the orientation of an array of nozzles belonging to a print head is parallel to the slow scan orientation.
  • the i-orientation shall by convention mean the fast scan orientation and the j -orientation the slow scan orientation, although the current invention also works using a different convention.
  • the positions on the substrate where a droplet can be printed form an addressable grid of pixels.
  • the addressable grid of pixels is also called a raster, comprising raster lines that are parallel to a fast scan orientation.
  • a raster line shall refer to a line of pixels parallel to the fast scan orientation, and is indexed by means of a [i] coordinate.
  • a column of pixels is indexed by a [j] coordinate.
  • a pixel in an addressable grid is hence addressed using a set of [i,j] coordinates.
  • each nozzle of a print head can be indexed by means of a nozzle index nozzlelndex that ranges from 1 to nbrNozzles.
  • the shortest distance between two nozzles 130 along the slow scan orientation 140 is called the nozzle pitch 110 and is indicated by the parameter nozzlePitch.
  • the length 150 of a nozzle array can be expressed as a multiple of the length of slowScanPitch and is indicated by a parameter headSize.
  • a set of pixels that can be addressed by the nozzles of a print head during one movement along the fast scan orientation is called a swath.
  • the size of a swath is measured in an orientation parallel to the slow scan orientation is called swathSize.
  • the swath size is equal to or larger than the image of a document so that all the pixels of said document can be printed in one swath.
  • the nozzles 430 of an array may be staggered for constructive reasons along two or more columns (460, 461) .
  • the nozzle pitch 410 is defined as the shortest distance between two lines perpendicular to the slow scan orientation and going through the centers of the staggered nozzles.
  • the timing of the firing of droplets from nozzles belonging to different rows is preferably adjusted so that the pixels belonging to a same line parallel to the slow scan orientation in the image of a document also land on the same line parallel to the slow scan orientation on the printed image. By adjusting the timing this way, the processing to prepare the signals for the nozzles can be the same as if all the nozzles were virtually on one and the same line.
  • each head is used but a sub-assembly of preferably two heads that are mounted back to back.
  • Each individual print head comprises one row of 318 nozzles having a nozzle pitch of 169 micrometer, which corresponds to 59,2 nozzles per cm or 150,3 nozzles per inch.
  • the transducers in the print heads are of the electro-mechanical type and are built using piezoelectric materials.
  • a second head 520 is displaced over exactly half a nozzle pitch 541 with regard to a first head 521, so that a sub-assembly 530 is obtained that can print at twice the resolution of the original heads. So the sub-assembly is effectively equivalent to a single print head having two rows of 318 nozzles and having a nozzle pitch of 85,5 micrometer, which corresponds to 118,4 nozzles per cm or 300,6 nozzle per inch.
  • the swath size is increased by mounting a plurality of sub- assemblies 530 of print heads in a staggered fashion to form a print head assembly 500.
  • the sub-assemblies are displaced along a slow scan orientation over a distance so that any of the nozzle positions is essentially spaced along a slow scan orientation at a distance from any other nozzle position over a distance that is an integer number times the nozzle pitch 540. This is also indicated by the grid of dotted lines parallel to a fast scan orientation in Figure 5.
  • the sub-assemblies 530 may or may not have overlapping nozzles.
  • the timing of the firing of droplets from nozzles belonging to different print head sub-assemblies is preferably adjusted so that the pixels belonging to a same line parallel to the slow scan orientation in the image of a document also land on the same line parallel to the slow scan orientation on the printed image. By adjusting the timing this way, the processing to prepare the signals for the nozzles can be the same as if all the nozzles were virtually on the same line.
  • a set of four print head assemblies are used that print with different inks, such as with cyan, magenta, yellow and black inks or with inks having a different tint and or density.
  • the nozzles belonging to a different print head assembly are lined up along the fast scan orientation so that nozzles belonging to a different head assembly but having the same nozzle index print on the same raster line during the same swath.
  • the printing resolution in the fast scan orientation - represented by a parameter fastScanResolution - is proportional with the firing frequency of the nozzles represented by a parameter firingFrequency.
  • the parameter firingFrequency stands for the time rate at which ink droplets can be ejected by a nozzle.
  • the fast scan resolution fastScanResolution is dictated by the ratio of the firing frequency firingFrequency divided by the velocity fastScanVelocity in the fast scan direction:
  • fastScanResolution FiringFrequency/fastScanVelocity
  • a higher spatial resolution enables the rendering of finer image details and hence generally enables to achieve a higher image quality.
  • printer commands are generated from a data processing system 600 such as a computer.
  • the data processing system 600 comprises a network connection means 621, a central processing unit 622 and memory means 623, which are all connected through a computer bus 624.
  • the data processing system 600 typically also has a computer human interface 630, 631 for inputting data and a computer human interface 640 for outputting data.
  • the computer program code is stored on a computer readable medium such as a mass storage device 626 or a portable data carrier 650, which is read by means of a portable data carrier reading means 625.
  • a substrate transport motor 740, a curing station 750 and actuators of print heads 730 are controlled by a printer controller 700.
  • Printer commands 710 are received by a buffer memory 701. These printer commands contain controller information, which is sent to controlling means 705 and image data that is sent to an image buffer 702.
  • An image processor 703 can operate on the image data in the image buffer 702.
  • the image processor 703 has access to one or more threshold masks 711 for the purpose of halftoning.
  • the controlling means 705 controls a substrate transport driver 707 that drives said substrate transport motor 740.
  • the controlling means 705 also controls a curing station driver 706 that drives a curing station 750.
  • the information in the image buffer 702 can be used to drive the actuators of the print heads 730 by means of a print head driver 704.
  • the image processor 703 in the printer controller 700 comprises a halftoning processor for converting a continuous tone image into a halftone image .
  • said halftoning processor uses one or more threshold masks 711.
  • Figure 8 shows a first embodiment of a halftone processor.
  • a continuous tone image 820 comprising pixels P[i,j] 860 represented with 8 unsigned bits per pixel is converted into a halftone image 830 comprising halftoned pixels H[i,j] 850 represented with 1 (unsigned) bit per pixel.
  • This is achieved by using a threshold mask 810 comprising threshold values S[i,j] 880 that are preferably represented on a scale from 1 to 255.
  • An address generator 840 generates addresses [i,j] of the halftone image 830. These addresses [i,j] are converted by an address converter 870 into converted addresses [i',j']. The address conversion may comprise position, scaling and rotation coordinate transformations.
  • the address [i',j'] is used to address a continuous tone pixel P[i',j'l 860 of an image 820.
  • the value of the continuous tone pixel is P[i'j'] is multiplied by an integer factor K in a multiplier 861 to obtain a multiplied value K*P[i'j'].
  • K is preferably equal to one.
  • the addresses [i,j] are also converted by a modulo converter 871 into modulo converted addresses [i'',j''] .
  • These modulo converted addresses [i'',j''] address threshold values S[i'',j''] 880 in a threshold mask 810 comprising iSize by jSize thresholds.
  • i' ' i modulo iSize
  • j ' ' j modulo jSize
  • the conversion of the continuous tone pixel P[i',j'] 860 into a halftone pixel H[i,j] 850 is obtained by adding to the multiplied continuous tone pixel K*P[i',j'] 860 a threshold value S[i'',j''] 880, and quantizing the result K*P[i ' , j ' ] +S [i ' ' , j ' ' ] by a quantizer 891.
  • the quantization comprises obtaining the most significant bit of a nine bit word that represents said sum P[i ' , j ' ] +S [i ' ' , j ' ' ] .
  • FIG. 9 shows a one dimensional cross section of an image having pixels P[i',j'] and of a threshold mask having threshold values S[i'',j''].
  • the values of the image and threshold mask are drawn as continuous lines, although in reality the image and the threshold mask values are discrete samples in space having discrete values.
  • the values of the threshold mask that are shown are representative for periodical screening as opposed to for frequency modulation screening. It is clear however, to the person skilled in the art that said halftoning processor also works when a threshold mask is used that has been specifically calculated for the purpose of frequency modulation halftoning as in one of the other embodiments of the current invention.
  • FIG. 9 corresponds to a halftoning process for a printer that can render two density values that are binary coded as 0 and 1. Variations of the method exist that also work for s printer than can render more than two density levels .
  • Figure 10 shows a first embodiment of a halftone processor for a printer that can render not two but four halftone values H[i,j] having values 0 , 1 , 2 or 3.
  • K is preferably set to a value of three and the quantizer is preferably programmed to calculate the two most significant bits of a 10 bit word that represents the sum 3*P[i'j'] + S[i' ', j"] .
  • H[i,j] ( (OBlI.0000.0000)&(K*P[i, j]+S[i,j] ) )»8;
  • the multilevel screening reduces to a binary screening method according to a previous embodiment.
  • FIG. 11 Another embodiment of a multilevel halftoning processor is shown in Figure 11.
  • threshold masks not one but multiple threshold masks are used. Specifically referring to Figure 11, a set of three threshold masks are used corresponding to threshold values Sl[i",j"], S2[i",j”] and S3[i",j”] .
  • the value H[i,j] that is generated for a halftone pixel depends on the value of a pixel P[i,j] relative to the values in the three threshold values Sl[i",j"], S2[i",j”] and S3 [i '',j"] in the threshold masks .
  • H3 [ i ,j ] ( ( OBl.0000.0000 ) & (P [ i , j]+S3[ i ,j])) »8 ;
  • H[i,j] is calculated based on the possible outcomes of Hl[i,j], H2[i,j] and H3[i,j]:
  • a set of 5 threshold masks would be used and the following table could serve to convert a set of 5 intermediate binary halftone values into a multilevel halftone value.
  • a halftoning processor is implemented using software code that runs on a general purpose data processing system such a system shown in Figure 6.
  • a halftone processor is implemented in an embedded system, for example using ASIC or FPGA.
  • a first step 1210 consists of defining an integer size of the threshold mask that is to be calculated.
  • the threshold mask can be square or rectangular.
  • the threshold mask is preferably square and has a size of iMatrixSize in the i- orientation and jMatrixSize in the j -orientation.
  • the mask size is a power of 2. Good results can be obtained with a threshold matrix having a size of 256 by 256 thresholds.
  • the mask size is a prime number, for example 491.
  • two matrices are declared in a step 1220, both having a size of iMatrixSize by jMatrixSize.
  • the matrix rasterMountain contains preferably floating point values. It is preferably initialized with small random numbers, such as numbers produced by a white noise random generator. The numbers can for example be in a range between 0.0 and 10. OE-2.
  • the matrix orderNumberMatrix contains preferably integer values that are preferably initialized to zero, although other initial values are also possible.
  • the domain of the function is preferably limited and is preferably rectangular or elliptical.
  • FIG. 14A, 14B, 14C and 14D Different embodiments of a domain are shown in Figure 14A, 14B, 14C and 14D.
  • the size of the domain is defined by iDomainSize 1401 along a first axis parallel to the i-orientation and jDomainSize 1402 along a second axis parallel to a j -orientation.
  • the ellipse is rotated over an angle alpha 1403 with regard to the i-orientation and j -orientation. According to the preferred embodiment, this angle alpha is equal to zero degrees and the long and short axes of the ellipse are parallel with the i-orientation and j -orientation.
  • iDomainSize is odd and larger than iMatrixSize and preferably, jDomainSize is also odd and larger than jMatrixSize, for example, double the corresponding matrix size.
  • the size of the domain is defined by iDomainSize 1401 along a first axis parallel to a i-orientation of the rectangle and jDomainSize 1402 along a second axis parallel to a j -orientation of the rectangle.
  • the rectangle is rotated over an angle alpha 1403 with regard to the i-orientation and j -orientation. According to the preferred embodiment, this angle alpha is equal to zero degrees and the sides of the rectangle are parallel with the i-orientation and j -orientation.
  • iDomainSize is odd and larger than iMatrixSize and preferably, jDomainSize is odd and larger than jMatrixSize, for example, double the corresponding matrix size plus one.
  • the single valued function is symmetrical with regard to the two main axes of the domain.
  • Figure 16 shows an example of such a function.
  • profiles that can be used to define a single valued function are:
  • a rectangular matrix tableFunction with iDomainSize by jDomainSize floating point elements is declared and initialized in a step 1240 ( Figure 12) .
  • the value of an element 1310 is set to a value that is representative for the volume defined by the rectangular area 1320, the four ribs 1330 and the upper surface 1340.
  • the value of the integral is approximated using one of the techniques as known in the art under the name of numerical integration or quadrature.
  • a next step 1250 comprises defining and declaring a loop counter loopCntr, which is preferably initialized at a value 1.
  • This loop counter can be used to assign order numbers to elements of the matrix orderNumberMatrix.
  • the lowest order number is preferably 1, the highest order number preferably iMatrixSize*jMatrixSize.
  • one interpretation of an order number at a given position in the matrix orderNumberMatrix is that it indicates at which point in the sequence of adding halftone dots for increasing the density of a tint from zero to maximum a halftone dot is added at the corresponding position of the order number.
  • a first step 1251 of the main loop comprises determining a set of coordinates [iMin,jMin] of the element in the matrix rasterMountain that has a global minimum value minRasterMountain and additionally of which the corresponding element in orderNumberMatrix has not received an order number. Determining such a global minimum is preferably done using one of the numerical sorting techniques as they are known in the art of numerical computing.
  • a second step 1252 of the main loop comprises assigning to the element with position (iMin,jMin) in the matrix orderNumberMatrix the value of loop counter loopCntr.
  • a third step 1253 of the main loop comprises shifting all the elements of the matrix tableFunction along the i and j coordinates by a an offset so that the center of the matrix tableFunction corresponds with the position (iMin,jMin) of the global minimum in rasterMountain. As part of the same step 1253, these elements are added one to one to the elements of the matrix rasterMountain.
  • a third step 1254 of the main loop comprises incrementing the loop counter loopCntr of the main loop by one .
  • the value of the loop counter loopCntr is checked. If the loop counter loopCntr is equal to preferably iMatrixSize*jMatrixSize, the main loop is exited, otherwise the steps one to four of the main loop are repeated.
  • the method according to above description results in a matrix orderNumberMatrix with values ranging from 1 to iMatrixSize*jMatrixSize .
  • a step 1260 comprises rescaling the values of the elements in the matrix orderNumberMatrix into a tonal range from 1 to preferably 1 to 65,536.
  • a step 1270 comprises mapping the values of the elements in the matrix orderNumberMatrix that are represented with first, higher number of bits to values represented with a second lower number of bits using a tone curve.
  • said first higher number of bits is 16 and said second lower number of bits is 8.
  • step 1270 is a threshold mask that can be used for frequency modulation halftoning, for example by using the scheme that is shown in Figure 8.
  • the method that has just been presented is computationally effective: no transformation between the frequency and space domain are required, no filtering and no plural iterations are required as in the prior art methods.
  • the method produces halftone distributions having blue noise characteristics, because the addition of the single valued function to an element of the rasterMountain has the effect that the position of a halftone dot that is added to increase the average density of a halftone dot distribution is not nearby the position of any of the dots to render a lower average density.
  • the method also comprises a step to create from the halftoned image (830) a printing plate precursor.
  • a printing plate precursor is an imaging material that can be used as a printing plate after one or more treatment steps that include image-wise exposure and possibly processing.
  • the method also comprises a step to print the halftoned image (830) on a substrate using an inkjet printing apparatus .
  • the above description represents a method to calculate a threshold matrix for the purpose of frequency modulation halftoning that requires fewer calculations than the methods in the prior art that are based on using void and cluster filtering and transformations to the Fourier domain. Instead, the method according to the invention requires for the calculation of the position of each halftone dot only the addition of the elements of a first matrix to the elements of a second matrix, and the result of a sorting algorithm for which efficient implementations exist.
  • the above method also results in dot profiles that have desirable blue noise characteristics because the effect of the addition of the matrix tableFunction to the matrix rasterMountain is designed so that the position of a next minimum value in the matrix rasterMountain is moved away from the position of the previous minimum in the matrix rasterMountain.
  • the position of a next halftone dot is preferably placed at a distance from the position of a previous halftone dot.
  • the position of next halftone dot is preferably placed at a distance from the position of the previous but also at a distance from the positions of all previously placed halftone dots.
  • the single valued function creates a repulsive force field that inhibits the placement of a next halftone dot in the immediate area 30 of previously placed halftone dots.
  • the method according to the current invention is capable of producing dot distributions of which the power spectrum has a blue noise characteristic. abstraction of said method to calculate of a single threshold mask
  • a threshold mask for frequency modulation halftoning can be summarized as: - first initialize a force field having the same domain as the periodical expansion of the domain of said threshold mask;
  • equivalent results are obtained when the two dimensional domain of the single valued function is not limited or if it has a shape that is different from the rectangular or elliptical shape in the preferred embodiments.
  • Equivalent results are also obtained when a single valued function is used that does not have a global maximum but a minimum at or near its origin. In that case the step 1251 in Figure 12 is preferably replaced by finding a maximum, rather than a minimum in the matrix rasterMountain.
  • threshold masks for color printing
  • the printing of color images is usually performed by halftoning and printing of a sub- images with cyan, magenta, yellow and black ink.
  • a set of plural and distinct threshold masks are calculated using one of the previously described methods to calculate a single threshold mask.
  • the multiple threshold masks are calculated in a way that the correlation between the dot distributions produced by said threshold masks is constraint.
  • Figure 20 shows a main loop 2000 that is a further development of the main loop 1200 in Figure 12.
  • rasterMountainC rasterMountainC
  • rasterMountainM rasterMountainM
  • rasterMountainY rasterMountainY
  • the elements of the matrices orderNumberMatrixC, orderNumberMatrixM, and orderNumberMatrixY are calculated in the same main loop 2000.
  • a first step 2001 in said main loop comprises determining a set of coordinates [iMin,jMin] of the element in the matrix rasterMountainC that has a global minimum value minRasterMountain and additionally of which the corresponding element in orderNumberMatrixC, orderNumberMatrixM and orderNumberMatrixY has not received an order number .
  • a second step 2002 in said main loop comprises assigning to the element with position [iMin,jMin] in the matrix orderNumberMatrixC the value of loop counter loopCntr.
  • a third step 2003 of the main loop comprises shifting all the elements of the matrix tableFunction along the i and j coordinates by a an offset so that the center of the matrix tableFunction corresponds with the position [iMin,jMin] of the global minimum in rasterMountainC.
  • these elements are added one to one to the elements of the matrix rasterMountainC, rasterMountainM, and rasterMountainY.
  • the effect of adding the values of the elements of the matrix tableFunction to rasterMountainC, to rasterMountainC, and to rasterMountainC is that, the position of a next minimum in any of the matrices rasterMountainC, rasterMountainM, and rasterMountainY is moved away from the position [iMin,jMin] that was found in rasterMountainC.
  • next steps 2011, 2012 and 2013 are similar to the steps 2001, 2002 and 2003, and the same is true for the steps 2021, 2022 and 2023.
  • the main loop 2000 is executed iMatrixSize*jMatrixSize/3 times. At that point iMatrixSize*jMatrixSize/3 times orderNumberMatrix elements in the matrices rasterMountainC, rasterMountainM, rasterMountainY have received an order number.
  • the positions of these elements in the three matrices are complementary, in that all the elements have different positions, and no position of such an element is common to a position of another element in another of said two matrices .
  • a general condition is that at the end of a main loop 2000, 2100 the total number of elements that have received an order number is preferably equal to iMatrixSize*jMatrixSize.
  • Figure 22 shows a first preferred embodiment according to the current invention. We use the notation C, M, Y and K to refer to the respective order number matrices .
  • the elements of a set of four order number matrices C, M, Y, K receive order numbers from 1 to iMatrixSize*jMatrixSize/4 in one combined loop.
  • the elements of a first set of two order number matrices C, M receive order numbers from 1+iMatrixSize*jMatrixSize/4 to iMatrixSize*jMatrixSize/2 in a combined loop. Also in said second step, the elements of a second set of two order number matrices Y, K receive order numbers from 1+iMatrixSize*jMatrixSize/4 to iMatrixSize*jMatrixSize/2 in a third combined loop.
  • Figure 23 shows another preferred embodiment.
  • the elements of a set of three order number matrices C, M, Y receive order numbers from 1 to iMatrixSize*jMatrixSize/3 in one combined loop.
  • a fourth matrix K receives order numbers from 1 to iMatrixSize*jMatrixSize/3 in a separate loop.
  • the elements of a first set of two order number matrices C, M receive order numbers from 1+iMatrixSize*jMatrixSize/3 to iMatrixSize*jMatrixSize/2 in a combined loop. Also in said second step, the elements of a second set of two order number matrices Y, K receive order numbers from l+iMatrixSize*jMatrixSize/3 to iMatrixSize*jMatrixSize/2 in a third combined loop.
  • each one of the order number matrices C, M, Y and K receive order numbers from l+iMatrixSize*jMatrixSize/2 to iMatrixSize*jMatrixSize in a separate loops .
  • the above describe method to calculate a set of threshold masks for frequency modulation halftoning is based on the previously described method for calculating a single threshold mask using a force field, but wherein, at least over part of the range, the assignment of order numbers in the different threshold masks is based on the same force field.
  • a set of children threshold masks is derived from a single threshold mask or order number matrix that serves as a parent matrix.
  • Figure 17 shows an example of a small threshold mask that serves as parent matrix.
  • the size of the matrix along the i- orientation iMatrixSize 1702 and the size of the matrix along the i- orientation iMatrixSize 1702 are both equal to three so the parent matrix comprises 9 threshold values.
  • These threshold values can be ordered from smallest to largest and be given an order number. Said order number can be assigned to every position in the threshold matrix.
  • the values 1, 2, 3,... 9 in the indicate the order number of a threshold in the series of threshold values in the parent matrix.
  • Table 1 The relation between the order number and the position of the corresponding threshold is also conveniently represented in Table 1 below.
  • the order numbers of a series of threshold values are directly obtained from the matrix orderNumberMatrix that is calculated in a main loop 1200 in Figure 12B.
  • a printing device is capable of reproducing 5 density levels (or equivalent dot areas or droplet volumes) .
  • said levels can be coded as level 0, level 1, level 2, level 3 and level 4.
  • the method works just as well with any other number of levels.
  • a number of printable levels is equal to a parameter nbrLevels ranging from level 0 to level nbrLevels-1.
  • a two-dimensional table can be made, called order number table that comprises a level index x and a position index y.
  • a level index x of said table corresponds to a printable level of a printer.
  • the order of the levels is arranged so that a larger level index preferably corresponds to printing with a higher density.
  • a position index y of said table corresponds to a position in the parent matrix.
  • the order of the positions is arranged so that a larger position index preferably corresponds with a larger threshold value at the corresponding position in the threshold mask.
  • Table 2 below shows an example that corresponds with the threshold mask shown in Figure 17 and in which the level index corresponds with a row index and a position index corresponds with a column index .
  • the method comprises a step in which said function is used to populate said order number table.
  • the method comprises a step in which each function value F(x,y) in Table 3 is assigned a unique order index o that corresponds with the order number of said function value F(x,y) in said series of function values .
  • the position in Table 3 corresponding with level 0 and position [0,2] receives the smallest function value and is assigned an order index 1.
  • the position in Table 2 corresponding with level 0 and position [0,1] receives the second smallest function value and is assigned the order index 2. This process of assigning order numbers is repeated until all the positions in the order number table have received an order index.
  • the result of the above step is an order number table filled with order indices as shown in Table 4 below.
  • the method comprises a step to transform the elements of the order number table obtained in a previous step into a threshold masks that can be used for multilevel halftoning using a scheme as , for example, the one explained by means of Figure 11 in the current application.
  • This transformation comprises two sub-steps.
  • the order numbers in said order number matrix are transformed into an appropriate range of threshold values.
  • the range of order numbers in Table 4 goes from 1 to 45.
  • a range of threshold values is needed preferably having a range from 1 to 255.
  • the transformation of the range can be achieved in different ways.
  • the elements of the order number matrix are simply multiplied by an appropriate scale factor.
  • an appropriate scale factor would be 255/45.
  • the multiplication of all the elements in Table 4 by said multiplication factor yields a new order number table that is shown in Table 5.
  • a look up table approach is used to map the range of values in the order number table to the appropriate range of values for the purpose of digital halftoning.
  • Figure 19 maps the range of the values of the elements of an order number table - represented on an axis 1910 and having a maximum 1911 - to a range of threshold values represented on a vertical axis 1920 and having a maximum 1921 using a mapping curve 1930.
  • the mapping curve 1930 in Figure 19 is not a linear mapping curve, but is designed to achieve a designed gradation of the halftoned image in the final printed result. A method to design such a curve is for example explained in US 5,818,604.
  • the result of this step is a transformed order number table.
  • the method comprises a step to separate said transformed order number table into a set of nbrLevels threshold masks. This is achieved by assigning every element of the transformed order number table to an element having a position [i,j] in a threshold mask, wherein the position [i,j] corresponds to the position index of said element in said transformed order number table, and wherein said threshold mask corresponds to the level index of said element in said transformed order number table .
  • Figure 18A to Figure 18E shows a set of threshold masks that are obtained by separating Table 5.
  • the result of this step is a set of nbrLevels threshold masks that can be used for multilevel halftoning using a method that was explained by means of Figure 11.
  • the method according to the current invention enables to control the balance between increasing a number of halftone dots or to increase the density of individual halftone dots to increase the density of a printed tint. abstraction

Landscapes

  • Physics & Mathematics (AREA)
  • Discrete Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Image Processing (AREA)
  • Facsimile Image Signal Circuits (AREA)

Abstract

L'invention concerne un procédé permettant de calculer un ensemble de masques de seuil pour similigravure numérique à colorants multiples. Ce procédé présente une haute efficacité de calcul et produit une distribution de points dont le spectre de puissance présente les caractéristiques d'un bruit bleu. Ce procédé fait appel à une fonction uniforme, présentant un domaine bidimensionnel limité. La fonction uniforme présente un maximum situé à, ou à proximité de son origine et décroît de manière monotone en direction des limites du domaine bidimensionnel limité. La distribution des points correspondant aux différents colorants est optimisée entre les différents points afin de réduire la visibilité de la granulation. Ce procédé permet de commander des imprimantes capables de rendre deux niveaux de densité par pixel ou davantage.
PCT/EP2006/063461 2005-06-24 2006-06-22 Trame de similigravure a modulation de frequence multiniveau et procede permettant de produire cette trame WO2006136600A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2008517509A JP2008544667A (ja) 2005-06-24 2006-06-22 多数レベル周波数変調ハーフトーンスクリーンとその製法

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
EP05105659.6 2005-06-24
EP05105659A EP1737211A1 (fr) 2005-06-24 2005-06-24 Trame de demi-teintes à modulation de fréquence et à plusieurs niveaux et son procédé de fabrication
US69786405P 2005-07-08 2005-07-08
US60/697,864 2005-07-08

Publications (1)

Publication Number Publication Date
WO2006136600A1 true WO2006136600A1 (fr) 2006-12-28

Family

ID=37136786

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/EP2006/063461 WO2006136600A1 (fr) 2005-06-24 2006-06-22 Trame de similigravure a modulation de frequence multiniveau et procede permettant de produire cette trame

Country Status (2)

Country Link
JP (1) JP2008544667A (fr)
WO (1) WO2006136600A1 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8934142B2 (en) 2013-04-30 2015-01-13 Hewlett-Packard Indigo B.V. Halftone screen

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0578110A2 (fr) * 1992-06-29 1994-01-12 Eastman Kodak Company Appareil pour la génération d'images en demi-teintes à plusieurs niveaux
EP0719031A2 (fr) * 1994-12-22 1996-06-26 Eastman Kodak Company Procédé pour minimiser l'erreur de quantification dans des fonctions de transfert de demi-teintes à plusieurs niveaux
US5689586A (en) * 1994-03-02 1997-11-18 Hewlett Packard Company Methods to print N-tone images with multi-leveling techniques
US5729663A (en) * 1995-12-07 1998-03-17 Xerox Corporation Method and apparatus for gray screening
US6068361A (en) * 1997-10-30 2000-05-30 Mantell; David A. Method and apparatus for multiple drop error diffusion in a liquid ink printer

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0578110A2 (fr) * 1992-06-29 1994-01-12 Eastman Kodak Company Appareil pour la génération d'images en demi-teintes à plusieurs niveaux
US5689586A (en) * 1994-03-02 1997-11-18 Hewlett Packard Company Methods to print N-tone images with multi-leveling techniques
EP0719031A2 (fr) * 1994-12-22 1996-06-26 Eastman Kodak Company Procédé pour minimiser l'erreur de quantification dans des fonctions de transfert de demi-teintes à plusieurs niveaux
US5729663A (en) * 1995-12-07 1998-03-17 Xerox Corporation Method and apparatus for gray screening
US6068361A (en) * 1997-10-30 2000-05-30 Mantell; David A. Method and apparatus for multiple drop error diffusion in a liquid ink printer

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8934142B2 (en) 2013-04-30 2015-01-13 Hewlett-Packard Indigo B.V. Halftone screen

Also Published As

Publication number Publication date
JP2008544667A (ja) 2008-12-04

Similar Documents

Publication Publication Date Title
US7936479B2 (en) Image processing method for reducing imaging artifacts
US10442211B2 (en) Dual pass uniformity printing compensation mechanism
US7821672B2 (en) Image recording method and image recording apparatus
US6962400B2 (en) Sub-dot phase modulation for computer to plate inkjet system
JP2006042332A (ja) 階調依存型マルチ周波数ハーフトーンスクリーニングのためのシステム及び方法
EP1737211A1 (fr) Trame de demi-teintes à modulation de fréquence et à plusieurs niveaux et son procédé de fabrication
US7342685B2 (en) Constraint correlation for computer to plate inkjet system
US8169666B2 (en) Threshold matrix generation method, halftone image generation method, and halftone image generating apparatus
WO2007088154A1 (fr) Courbe de séparation de niveau de conception conçue pour des épreuves simili
US7440141B2 (en) Method for generating a dither mask
WO2002042079A1 (fr) Generateur de masque et motifs de masque d'image
AU2002227320A1 (en) Mask generator and image mask patterns
WO2006136600A1 (fr) Trame de similigravure a modulation de frequence multiniveau et procede permettant de produire cette trame
WO2006136599A1 (fr) Ensemble de trames de similigravure a modulation de frequence et procede permettant de produire ces trames
US20060279788A1 (en) Automatic generation of supercell halftoning threshold arrays for high addressability devices
EP1401190B1 (fr) Modulation de la phase des sous-points pour les systèmes d'ordinateur à plaque par jet d'encre
WO2013054319A1 (fr) Génération de données pour commander l'éjection de gouttes d'encre
US6049395A (en) Method and system for achieving enhanced gray levels in a screen cell array
EP1816850A1 (fr) Tirage d' épreuves en demi.teinte avec une batterie de courbes adaptées
US20060274338A1 (en) Method and system for inter-channel clock frequency selection with periodic halftoning
EP1608146B1 (fr) Procédé pour la génération d'une matrice de tremblement
JP2024006471A (ja) 画像処理装置、閾値マトリクス生成装置、それらの方法およびプログラム

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application
WWE Wipo information: entry into national phase

Ref document number: 2008517509

Country of ref document: JP

WWE Wipo information: entry into national phase

Ref document number: 200680022763.X

Country of ref document: CN

Ref document number: 5924/CHENP/2007

Country of ref document: IN

NENP Non-entry into the national phase

Ref country code: DE

WWW Wipo information: withdrawn in national office

Ref document number: DE

122 Ep: pct application non-entry in european phase

Ref document number: 06763854

Country of ref document: EP

Kind code of ref document: A1