CN1248752A - Color semi-color tone micropoint coding method for high density information storage - Google Patents

Color semi-color tone micropoint coding method for high density information storage Download PDF

Info

Publication number
CN1248752A
CN1248752A CN 98120704 CN98120704A CN1248752A CN 1248752 A CN1248752 A CN 1248752A CN 98120704 CN98120704 CN 98120704 CN 98120704 A CN98120704 A CN 98120704A CN 1248752 A CN1248752 A CN 1248752A
Authority
CN
China
Prior art keywords
data
unit
color
cell
data cell
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
Application number
CN 98120704
Other languages
Chinese (zh)
Inventor
孙祖洪
吴汤渝生
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Individual
Original Assignee
Individual
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
Application filed by Individual filed Critical Individual
Priority to CN 98120704 priority Critical patent/CN1248752A/en
Publication of CN1248752A publication Critical patent/CN1248752A/en
Pending legal-status Critical Current

Links

Images

Landscapes

  • Record Information Processing For Printing (AREA)

Abstract

The present invention relates to a halftone micropoint coding method for storing large volume of data. Said micropoint is produced on a base lining by means of high-resolution colour laser or ink-jet printer or lithographic press. It adopts binary data coded data unit, and the combination of data units is fermed into a data block. Said invention uses external data unit to enclose data block, and to define data unit space, data block orientation and data unit reflectivity in the course of scanning and data reconstittuion. In said method, every byte possesses a parity check bit, and possesses the functions of error detection and correction.

Description

Be used for little some coding method of color halftoning of high density information storage
The present invention relates to non-volatile photoelectricity ROM (read-only memory), comprise the method for printing the storage of promising high density information thereon and serving as a contrast and form these symbols with the base of the two-dimensional array of the little point symbol of shadow tone colour (half-tone color) of long number binary data coding.
It is well-known to have increased the black and white high density two-dimentional bar code symbol that is stored in quantity of information in the one-dimensional bar code printed symbols.At U.S. Patent number 5,304,786 and U.S. Patent number 5,504,322 patents in the example of this two-dimensional bar is disclosed.The resolution of printer and scanner, the bar code symbol that is printed to straight (alignment) and during scanning and data reconstruction error rate limited the increase of two-dimentional black and white bar code density.
In production has the miniature label technique of color bar code, also used multiple colour, as at U.S. Patent number 5,129,974 and U.S. Patent number 5,118,369 in describe.Yet the system of described type typically is mainly used in manufacture process sign and adds little label for the small semiconductor integrated circuit.
At U.S. Patent number 4,824, disclose in 144,4,846,503 and 4,889,367 and utilized the coloud coding increase to be stored in quantity of information on the label (label).In order to increase the Data Loading capacity of data array, as at U.S. Patent number 4,488, that describes in 679 has developed the coding that uses the colored point of a plurality of high density.The broken colour coding of wherein describing (multi-color code) and read the printing that colour that code system utilizes five contrasts is used for yard.A plurality of points form the piece of representing distinct symbols.
U.S. Patent number 4,874,936 disclose a data array, and this data array comprises with the continuously arranged hexagon that is encoded of honeycomb-like pattern having at least two different optical characteristics information.U.S. Patent number 5,369,261 disclose density polychrome information coding system, this system adopts the super picture element that different colored subregions are arranged in it, this super picture element is piled up on another color region by each colored arrangement side by side or a color region and is arranged, like this, the information of a point contains the layer or the so much change in color of mixolimnion of piling up together.For every kind of color region nearly 64 kinds of intensity are arranged with overtrick (Superdot).Yet, the encoding scheme of Miao Shuing wherein, the encoded recording system with read code system and can only use by all are housed and become the color CRT screen of image information that tripack colour film is exposed, and with red, green and white light source reader blue color filter is realized, like this, whenever, next homogeneous beam irradiation once and by the catoptrical intensity of ccd sensor array detection is used to calculate every kind of pigment intensity and to being present in the information decoding in the super picture element.Super picture element encoding scheme can not be used color laser printer or ink-jet printer, because the printer pixel is only by cyan, and yellow, the half tone dot of magenta and black ink combination.
U.S. Patent number 5,426,289 disclose the coloud coding that the combination that utilizes row mark (column marks) and row phase mark (column phase marks) solves the problem read correct data, even be applied to for example flax product etc. of elastic substrate.In readout, obtain decoded data by the position that is combined in predetermined colored code table of searching 3 colour pase marks.Wherein disclosed encoding scheme solves the problem of reading authentic data from elastic substrate by the density of sacrificing information stores.
Be used for encoding high density numerical information on different base linings although each patent proposes the various technology of foregoing description, implement complexity and costliness.In addition, this encoding scheme can not be used for current general laser or ink-jet printer.
So just, need to provide a method and apparatus, it utilizes common color laser or the ink-jet printer form with point on a medium that high-density digital information is encoded, for the user reduces cost.
The invention provides a kind of coded data method in two-dimentional colored little point symbol mode of creating the storage numerical data, this method comprises the following steps: to transform letter-digital text and binary data file becomes binary data file; The described binary data file of encoding; Generation is based on the level and the vertical correcting unit of makeing mistakes of described coding binary unit; Produce boundary element in response to the described correcting unit of makeing mistakes; Generation contains the message file relevant for described data unit information; Be coupled described message file to printer.
The present invention also provides a kind of method of two-dimentional colour symbolism of numerical data of the expression storage of decoding, and this symbol comprises border and data cell, and this method may further comprise the steps: scan described symbol to produce the image data file of scanning; Read described border and data cell; The image file of the described scanning of verification wrong and proofread and correct described mistake; Produce binary data file; With generation coloured image and text from described binary data file.
The invention provides a kind of novel two-dimensional symensional symbol and be used for the coloured image of storage representation compression or a large amount of digital encoded data of text data file.This symbol can be used for for example serving as a contrast at a base foundes non-volatile ROM (read-only memory) on the paper, it can be used on the computer system as information stores parts.Can reach with technology of the present invention and to be listed on the area in square inches the nearly memory capacity of 3K bytes of memory device of storage.
Especially, basic data storage unit of the present invention is one and comprises by color printer, color inkjet printer, or the data cell of the little point of two-dimentional shadow tone of the color of the same race of offset press printing.The long number digital coding of representing by the colour of data cell of each data cell.Each has unique colored data cell is a unique combination of expression long number binary data.The number of the little point of each data cell color halftoning is determined by the resolution of color printer and scanner.Only reach the data cell of 6 point * 6 or 10mil * 10mil with the color laser printer of 600 of per inchs and colored CCD scanning.Each data cell is represented the long number binary data, and a plurality of adjacent data cell form a digital byte.Data block is to represent the combination of the data cell of whole data file with the two-dimensional array of a plurality of row and columns.Data block surround by a boundary element (boarder-cell) ring and wherein each data cell contain a predetermined colour or data type.This boundary element ring provides during the data decode orientation about data block, the expectation pixel value of interval between the data cell and different types of data.
The pixel value in C (cyan) Y (yellow) M (magenta) K (black) color space that sends to by appointment for the color data unit of each unique data type within the image file of printer defines.Specific long number binary data of each unique data types to express, and unique data type sum equals the sum of the various combination of multidigit binary data.By utilizing the basic printing ink of printer, that is, cyan, magenta, yellow, black reaches 8 unique data types and each data types to express 3 bit binary data in the present invention.By changing the pixel value in cyan or magenta part, shadow tone in the process printer and screen (screening) mechanism can produce contains nattierblue and light pinkish red colored data type.
Reading, or when decoding, with the measurement data unit to red, green or the high-resolution color ccd scanner of blu-ray reflection degree or the color data unit that the scanning of other reading device is printed.For data cell is decoded into binary digital data, each measured data cell is compared with the expectation pixel value of each data type in the pixel value of R (red), G (green), B (indigo plant) color space.Pixel value according to the expected data type that reads is defined in two threshold values in each R, G, the B chrominance component.The expectation pixel value of hope different types of data in the RGB color space has enough distances so that better noise resisting ability to be provided between the very big such threshold value of difference each other.The invention discloses a kind of algorithm, be decoded into the triad number than the pixel value that each data cell is measured by threshold ratio with pixel value and each colored composition.
Comprising in the data block of the present invention makes mistakes detects and bearing calibration.In each digital byte of representing by a plurality of horizontal adjacent data cell, comprise a parity bit, think that the detection that makes mistakes that is used for makeing mistakes the odd number position provides parity checking when reading.Each horizontal line is made up of data cell and horizontal Reed-Solomon (Reed-Selomon) verification unit in data block.Each vertical row also contains data cell and vertical Reed-Solomon verification unit in the data block, and it provides detects and proofread and correct the ability that a plurality of data cells are made mistakes in every row and every row.
The present invention also provides a kind of method and apparatus, utilizes computing machine and color laser printer, and color inkjet printer or offset press are created high density two dimension colour symbolism for the color document image of storage digital text file or compression.And by utilize computing machine and CCD chromoscan only or other reading device the symbol decoding that is printed is become text or becomes the coloured image of decompression.
The present invention also provides a kind of basis to be the scan image pixel value of maximum antinoise printer pixel value that is each chrominance section appointment and the measured data cell method and apparatus with color data cell encoding long number data, and according to the decoding algorithm to a plurality of threshold values of each chrominance component.The process that the present invention finishes coding/decoding like this is simpler and cheap but more accurate and reliable than current available coding/decoding color halftoning dot system.
For better understanding the present invention and other target and wherein further feature, please refer to following description and accompanying drawing;
Fig. 1 shows and utilizes the color laser printer cyan, magenta, and the coding schedule of eight data types of the color data unit of 3 bit digital data is represented in combination yellow and black ink, wherein imports pixel value for each data type appointment;
Fig. 2 is the standardization figure of pixel value in the RGB color space, and this pixel value is only to measure the light reflectance value with the red, green, blue coloured light line of eight kinds of data types of Fig. 1 coding schedule coding of acquisition with the CCD chromoscan;
Fig. 3 shows the improved coding schedule of eight unique data types of the color data unit of representing 3 bit digital data, it has utilized the nattierblue of color laser printer, light magenta, the combination of yellow and black, wherein nattierblue and light magenta are to specify less pixel value by shadow tone in the printer and the generation of screen mechanism for cyan and pinkish red chrominance section;
Fig. 4 is the standardization figure of pixel value in the RGB color space, and this pixel value is to measure the light reflectance value of the red, green, blue coloured light line of the eight kinds of data types using the improved coding schedule coding of Fig. 3 that obtain with the CCD color scanner;
Fig. 5 a shows a data byte by 3 adjacent color data cell encodings, and wherein first data cell is represented 3 highest significant position b 7b 6b 5, second data cell is represented a b 4b 3b 2, and the 3rd data cell is represented two least significant bit (LSB) b 1b 0With a parity bit P;
Fig. 5 b is two dimension 102 row * 102 column data pieces that are used for data information memory, wherein every row comprises that 96 data unit add that 6 horizontal check unit make mistakes and detect and proofread and correct to be used for level, and every row comprise that 96 data unit add that 6 vertical check unit detect and proofread and correct to be used for vertically makeing mistakes;
Fig. 6 shows the two-dimentional colour symbolism of a data block of expression, this data block is surrounded by the boundary element ring with tentation data, this ring upper left corner boundary element has unique data type, and it is directed that this boundary element is used for providing when reading, at interval with automatic calibration information;
Fig. 7 utilizes a color laser printer and different computer software module to create the process flow diagram of the detailed step of two-dimentional colour symbolism for the information stores of the image file of text and compression;
Fig. 8 utilizes a colored CCD to scan the process flow diagram of only with different computer software module two-dimentional colour symbolism being decoded and reproducing the detailed step of represented text or decompression color document image;
Fig. 9 is the example of word " Hello " with 15 color data cell encodings;
Figure 10 is detailed data cell coding process flow diagram;
Figure 11 is an example, and next the first line data piece with 96 data unit is data type 2 with data type 7 beginnings, and data type 1 or the like adds 6 horizontal check unit that produced by the Reed-Solomon error correcting code endways;
Figure 12 is by finding out the data cell deconv produces the horizontal check unit with the polynomial remainder of encoding example;
Figure 13 is the detail flowchart that utilizes Reed-Solomon error correcting code generation level and vertical check unit; With
Figure 14 is the more detailed block diagram that produces the boundary element ring of two-dimentional colour symbolism.
The print processing of widespread use comprises offset press, laser printer, and ink-jet, heat transmits and waits them to come down to halftoning apparatus and binary.Coloured image typically reproduces with four kinds of different chrominance sections: cyan, yellow, magenta and black.For each chrominance section, printer will or printed corresponding colored ink or skip on output point and print corresponding colored ink on the output point, and fortune depends on binary picture.The printing ink that is printed does not have shade.Therefore only can export cyan for each laser spots printer, yellow, one of 16 kinds of combinations of magenta and black ink.For with colour not of the same race and shadow representation continuous tone coloured image, comprise that the halftone technique of screen and error diffusion is used to guarantee that different local averages colored and shade show.The electronics screen is the half tone dot of varying dimensions and shape of a technology create to(for) each chrominance section, and it is by comparing in the color space pixel value within the halftone cell with a fixed number laser spots and a threshold matrix epiphase.Cyan in the combination, yellow, magenta is used for reproducing the continuous-tone image with different colored and shades with the size and dimension of the variation of black half tone dot.For most of laser printer, realize that the hardware and software of shadow tone function resides in printer.Usually be described with the image file of printer interface with the high level description language of industrial standard, Adobes Postscrip descriptive language for example, and the color pixel value of image is described with continuous tone in different color spaces.In the CYMK color space, used 32 bit binary data of each pixel are represented:
P={C, Y, M, K} be each colored composition C (cyan) wherein, Y (yellow), and M (magenta), K (black) can represent with 8 bit binary data with from 0 to 255 value.
Data cell 10 is information memory cells of the present invention.Each data cell 10 is made up of the color laser printing point of fixed number.Utilize the high-resolution color laser printer of 600 of per inchs, the data cell 10 of 10mil * 10mil comprises 6 point * 6 or total colored little points of 36 shadow tones.For long number binary data 12 being encoded, must define different discrete data unit with specific input pixel value with different color data unit.In order to represent 3 bit digital data, must define eight discrete data unit for each color data unit, the pixel value of each pixel equates in this tentation data unit.Fig. 1 shows for the input pixel of data cell by specifying C14, Y16, and M18 and eight different pixel values of the colored composition of K20 are represented the coding method of the digitalized data unit of 3 bit digital data 12.Fig. 1 also shows the image 51 that reproduces by color laser printer for the corresponding data unit.Type 0 data cell 22 has reproduced image black 10 and input pixel value P0={255, and 255,255,255}, it represents 3 bit binary data 00052.Class1 data cell 24 has reproduced image purple 26 and input pixel value P1={255, and 0,255,0}, it represents 3 bit binary data 00154.Type 2 data cells 28 have reproduced image magenta 30 and input pixel value P2={0, and 0,255,0}, it represents 3 bit binary data 01056.Type 3 data cells 32 have reproduced image red 34 and input pixel value P3={0, and 255,255,0}, it represents binary data 01158.Type 4 data cells 36 have reproduced image yellow 38 and input pixel value P4={0, and 255,0,0}, it represents binary data 10060.Type 5 data cells 40 have reproduced image cyan 42 and input pixel value P5={255, and 0,0,0}, it represents binary data 10162.Type 6 data cells 44 have reproduced image green 46 and input pixel value P6={255, and 255,0,0}, it represents binary data 11064.Type 7 data cells 48 have reproduced image white 50 and input pixel value P7={0, and 0,0,0}, it represents binary data 11166.In this encoding scheme, pixel value for each chrominance section is set at 255 or 0, this means to be used in (group's point screen or the random screen) independence of screen technology and independence of screen resolution and screen unit size in the printer, screen is 1 or 0 to whole data cell area scale-of-two pixel value afterwards.
For the 3 bit binary number digital coding data cells of decoding, with eight data unit of 600 high-resolution color ccd scanners of per inch scanning and printing.The CCD coloured image scanning of most of usefulness usually only have three red, the ccd image perception unit that green and blue optical filter is arranged in a group is scanned the ruddiness of image, the reflectance of green glow and blue light in order to measurement.In case scintigram picture, chromoscan only produce an image file, it has three color composition: R (red), G (green) and B (indigo plant).Each chrominance component pixel value is usually by representing for one 8 bit binary data of the measured reflectance of corresponding light source indication.Therefore, the image pixel binary value of scanning can be represented with one 24 place value:
S＝{r，g，b}。
By the pixel value sum of all pixels in the data cell is obtained the average scan image pixel value of a data unit divided by total number of data cell pixel.Because data cell type 748 is white, no printing ink, it should have red, and is green, the catoptrical maximal value number of blue ray.The reading of the average scan image of tentation data cell type 7 48 is S7={r7, g7, b 7, r7, g7, b 7Value is used for the reading of the average scan image pixel value of all other data types is carried out standardization, to obtain the relative data of each data cell type.Fig. 2 shows in R84 G86 B88 color space the figure to all data cell type standard scan image pixel values.Can find out that data cell type 748 does not have printing ink and it that maximal value S7={1.0,1.0,1.0}68 are arranged.Data cell type 0 22 has whole four kinds of printing ink and it should absorb the light of maximum.The reading of average scan image pixel value is S0={0,0, and 0}70.Blue or green printing ink has absorptive red light and reflection is green and the optical characteristics of blue light.Reading for the average scan image pixel value of data cell type 540 is S5={0.1,0.3, and 0.75}76.The optical characteristics of magenta ink is that it absorbs green light reflect red and blue light.Reading for data cell type 2 28 the average image pixel values is S2={0.75,0.1, and 0.2}82.Yellow ink absorbs blue light and reflect red and green glow.Reading for data cell type 4 36 average scan image pixel values is S4={1.0,1.0, and 0.2}78.Data cell type 332 has the combination of yellow and magenta ink, and it should absorb blue and green light rays reflection Red light.Reading for data cell type 3 32 the average image pixel values is S3={0.75,0.1, and 0.1}80.Data cell type 6 44 have blue or green and yellow ink combination it should absorb red with blue ray and reflection green light.The reading of average scan image pixel value is S6={0.1,0.3, and 0.2}74.Data cell Class1 24 have blue or green and magenta ink combination it should absorb redness and green light and reflection blue light.Reading for data cell Class1 24 average scan image pixel values is S1={0.1,0.1, and 0.2}72.According to the optical characteristics of each data cell type, the reading of the average scan image pixel value of data cell can be used for being decoded into 3 bit digital data.
Data cell type 2 28 only has magenta ink, and it should absorb green light rays and reflect red and blue light line.But the S2={0.75 that reads, 0.1,0.2}82 represents that the reflectance of blue light is low.This is because the effect of the higher density of the used magenta ink of printer, that is, the amount of per unit area magenta ink is too thick and cause the low reflection of blue light line.Similarly, data cell type 5 40 only has cyan ink, and it is answered absorptive red light and reflects green and blue light.But the S5={0.1 that reads, 0.3,0.75}76 represents that the concentration of cyan ink also is too high.As a result, the reflectance of the reflectance of the blue light of data cell Class1 24 and data cell type 6 44 green glows also is too low.In order to reduce the ink density of blue or green and magenta ink, during scanning and data decode, increase noise resisting ability, the green grass or young crops of all data type input picture files and the pixel value of magenta color composition are reduced to 123 from 255.Fig. 3 shows improved encoding scheme table.Data type 1 90 has reproduced image lavender 92, and it has input pixel value P1={123,0,123,0}.Data cell type 2 94 has reproduced image pink colour 96, and it has input pixel value P2={0,0,123,0}.Data cell type 4 102 with reproduced image yellow has an input pixel value P4={0,225,0,0}.Data type 5 106 with reproduced image blue light 108, it has an input pixel value P5={123,0,0,0}.Data cell type 6 110 with reproduced image green glow 112 have an input pixel value 123,255,0,0}.Set the input pixel value for specific chrominance section and equal 123, corresponding a halftone cell screen table that is undertaken by the shadow tone mechanism in the printer is to the pixel screen in the data unit.If this pixel value is greater than threshold value, scale-of-two output be 1 and printing ink be reproduced on this specified point.If this pixel value less than threshold value, does not have printing ink to be reproduced on this laser spots.Depend on used screen method (the vicissitudinous screen of putting in group's point screen of the size of the point that output changes or the output unit area at random of number) within the data cell only partial-pixel have that partial-pixel does not have printing ink within printing ink and the data cell.As a result, produce lighter color shade.In Fig. 3, data cell type 2 94 has input pixel value P2={0,0,123,0}.Reproduced image is pink colour 96 and the ink density that has reduced magenta ink.
With this improved encoding scheme, only scan the reproduced image of all eight kinds of different pieces of information cell types again with the high resolution CCD chromoscan of 600 of per inchs.Catch the output image file that is produced by scanner, calculate the average scan image pixel value of each data type, Fig. 4 shows in the RGB color space standardization figure for the average scan image pixel value of different pieces of information cell type.Comparable data cell type 7 48 carries out standardization, white cells 50, have red, green, the reflectance of blue light maximum.
S 0＝{0，0，0} 114
S 1＝{0.4，0.4，0.5} 116
S 2＝{0.9，0.4，0.6} 118
S 3＝{0.9，0.4，0.2} 120
S 4＝{0.9，0.9，0.2} 122
S 5＝{0.4，0.6，0.9} 124
S 6＝{0.4，0.6，0.2} 126
S 7＝{1.0，1.0，1.0} 128
As shown in Figure 4, the average scanning element value of eight different pieces of information cell types is separated from each other far in the RGB color space.When becoming 3 bit digital data, the image that transforms the data cell that is scanned in decode procedure, will provide noise resisting ability preferably.
When reading, with the scanning of 600 colored CCDs of high resolving power per inch only the two-dimentional colour symbolism that is printed of scanning and each data cell 8 unique data Class1 of average scanning element value and expectation pixel value 14 to 128 relatively, the data cell decoding is become 3 bit binary data.Following method of the present invention can be used to decoding:
The average pixel value of supposing a data unit is by following expression:
D＝{R，G，B}
R wherein, G, B are the standardized readings in each color space
Two threshold values of definition are as follows in each colored composition:
R 1＝0.2 R 2＝0.6
G 1＝0.2 G 2＝0.75
B 1＝0.35 B 2＝0.75
Following algorithm is used for D is decoded into a unique data type:
(1) if R 1＞R (possible data type is 0)
And G 1＞G (possible data type is 0)
And B 1＞B (possible data type is 0,3,4,6)
Data cell D is a type 0.
(2) if R 2＞R＞R 1(possible data type is 1,5,6)
And G 2＞G＞G 1(possible data type is 1,2,3,5,6)
And B 2＞B＞B 1(possible data type is 1,2)
Data cell D is that type is 1.
(3) if R＞R 2(possible data type is 2,3,4,7)
And G 2＞G＞G 1(possible data type is 1,2,3,5,6)
And B 2＞B＞B 1(possible data type is 1,2)
Data cell D is a type 2.
(4) if R＞R 2(possible data type is 2,3,4,7)
And G 2＞G＞G 1(possible data type is 1,2,3,5,6)
And B 1＞B (possible data type is 0,3,4,6)
Data cell D is a type 3.
(5) if R＞R 2(possible data type is 2,3,4,7)
And G＞G 2(possible data type is 4,7)
And B 1＞B (possible data type is 0,3,4,6)
Data cell D is a type 4.
(6) if R 2＞R＞R 1(possible data type is 1,5,6)
And G 2＞G＞G 1(possible data type is 1,2,3,5,6)
And B＞B 2(possible data type is 5,7)
Data cell D is a type 5.
(7) if R 2＞R＞R 1(possible data type is 1,5,6)
And G 2＞G＞G 1(possible data type is 1,2,3,5,6)
And B 1＞B (possible data type is 0,3,4,6)
Data cell D is a type 6.
(8) if R＞R 2(possible data type is 2,3,4,7)
And G＞G 2(possible data type is 4,7)
And B＞B 2(possible data type is 5,7)
Data cell D is a type 7.
Differing in any one color space between the expectation pixel value of the worst antinoise condition threshold and 8 unique data Class1 14 to 128 is 0.15.For example, for the expectation reading of data type 2 blue compositions be 0.6 and threshold value be G 1=0.2, G 2=0.75.Read data with blue composition is that the data cell between 0.2 and 0.75 is decoded into and is data type 2, also satisfies the needs of appointment in the algorithm in the red and green pixel value of this supposition.Therefore, be G for the worst tolerable noise of blue composition 2-0.6=0.15.Can see that far if the expectation average pixel value of eight unique data cell types separates each other in the RGB color space, that separates between the threshold value is also far.This step will be improved the noise resisting ability in the data cell decode procedure.
As shown in figures 1 and 3, for 3 bit binary number data coding schemes, each data cell represents that three horizontal adjacent data cell of one 3 bit digital data are used for representing that its unit, left side of a digital byte represents that the highest significant position right cell represents least significant bit (LSB).Fig. 5 a illustrates three highest significant position b that first data cell 130 is used for representing a byte 7b 6b 5132.Second data cell 134 is used for representing the position b of a byte 4b 3b 2136, the three data cells 138 are used for representing the least significant bit (LSB) b of a byte 1b 0Add a parity bit P 140.For even parity check, by under establish an equation and produce parity bit P 140:
P＝b 7+b 6+b 5+b 4+b 3+b 2+b 1+b 0。
Wherein+be xor operator b 7To b 0It is data bit and P is a parity bit.When reading, comprise that by each byte of verification the odd even of data bit and parity bit produces syndrome (syndrome) position.
S＝b 7+b 6+b 5+b 4+b 3+b 2+b 1+b 0+P。
Wherein+be xor operator
If there is the odd number position to make mistakes, the syndrome position of error byte equals logical one, otherwise equals 0.Therefore, can discern the Error Location of the byte of makeing mistakes of makeing mistakes the odd number position.This information is used to make mistakes and detects and the correcting scheme purposes.
Standard A SCII sign indicating number is used for a digital byte or 3 data unit any alphanumeric character being encoded.For example, according to the ASCII coding schedule, character " A " is encoded with 0,100 0001, and it can be shown by 3 data cell lists: 010 (data cell type 2), 000 (data cell type 0) then is that 010 (data cell type 2) has parity bit P=0.
The storage that the data block of forming with the form of two-dimensional matrix is used for a large amount of numerical informations of uniting by data cell.Fig. 5 b shows the data block of 102 row * 102 column data unit.For each row, 96 data unit 142 and 6 horizontal check unit 144 are arranged.96 horizontal data unit represent that 32 successive bytes and each byte have 3 data unit in the delegation.For example, first data byte is by data cell 150,152, and 154 constitute and the highest 3 significance bits of data cell 150 expressions.For each row, 96 data unit 146 and 6 vertical check unit 148 are arranged.The Reed-Solomon sign indicating number is used to produce six horizontal check unit and six vertical check unit, and six horizontal check unit have and detect and proofread and correct in each row maximum 3 and make mistakes at random.Six vertical check unit have maximum 3 of detecting and proofreading and correct in each row and make mistakes at random.The Reed-Solomon error correcting code is well-known in the prior art, is suitable for this most and uses, because the Reed-Solomon sign indicating number is towards symbol, each data cell is a symbol of 3 binary digits of expression.One makes mistake, two make mistakes or all three make mistakes and can cause that a data unit makes mistakes.Make mistakes to proofread and correct and mean the data cell of makeing mistakes with correct replacement.
Realize that from the Galois field of GF (8) its coding polynomial expression of Reed-Solomon sign indicating number is
g(X)＝(X+1)(X+α)(X+α 2)(X+α 3)(X+α 4)(X+α 5)
By finding out with the remainder generation horizontal check unit of horizontal data cell deconv with coding polynomial expression g (X) gained.Also produce the vertical check unit by finding out with the remainder of coding polynomial expression g (X) gained with vertical data unit deconv.When reading, the decoded binary data from the image file of CCD scanning is carried out the following step, be used for makeing mistakes of every row and detect and proofread and correct:
(1) by finding out with the whole line data unit deconv that comprises inspection unit, school six syndromes of each factor generation with generator polynomial expression (generator polymonimal).
(2) calculate the polynomial coefficient of mispointing device (error locator).When this step of beginning, the number of makeing mistakes is unknown.A kind of method that calculates the polynomial coefficient of mispointing device is at first to suppose a mistake.If find that this supposition is incorrect, the number of supposing of makeing mistakes increases by 1, or the like.
(3) find out the steady arm root of polynomial of makeing mistakes to determine the mispointing vector.
(4) calculate the polynomial logarithm value of mispointing device to obtain out mispointing.
(5) in a single day know the location vector of makeing mistakes, the syndrome equation becomes linear equation system, the value of wherein makeing mistakes is unknown.The calculating value of makeing mistakes.
(6) symbol that replaces the Error Location to locate with correct symbol.
Make mistakes by the odd number position if make mistakes and to cause.The parity checking of each byte also can be used to help the Error Location of finding out of step 4.Detect and correction repeats above-mentioned steps for makeing mistakes of each row on the vertical direction.Utilize above-mentioned disclosed data block design, comprise parity bit of each byte, 6 horizontal Reed-Solomon verification unit of every row, vertical Reed-Solomon verification unit with 6 has realized highly reliable data storage cell and can proofread and correct 3 data unit mistakes in every 3 data unit mistakes of row correction with at each row.
Utilization also comprises boundary element as shown in Figure 6 with the symbol of the present invention that the little point of the color halftoning of long number binary data coding is used for the high density information storage.The two-dimensional blocks of data 150 that comprises level and vertical check unit is at first surrounded by the ring of white data cell 152 (data cell type 7 48).The outside of white data unit loop 152 is boundary element rings 154.The boundary element ring has four angles 156,158,160 and 162.The data type 022 of the upper left corner data cell 156 of specified boundary unit loop (black data unit).Other three angular data unit 158,160 and 162 read with data decode during be used for orientation and to straight purposes.Since can be with any direction and angle scanning symbol, the data cell that the searching upper left corner has data type 0 will make decoding software can determine that the angle of image rotation is suitable aspect and decoded information.White data unit channel 152 between boundary element 154 and blocks of data elements 150 is used for separating blocks of data elements from the boundary element ring.The boundary element ring comprises that each unit has the data cell of known types.As an example, can specify the data type 1 of the horizontal data unit 164 that is close to the upper left corner 156, unit 166 is a data type 2, unit 168 is a data type 3, unit 170 is a data type 4, unit 172 is a data type 5, and unit 174 is a data type 6, and unit 176 is a data type 7 etc.Similarly, also can specify the vertical boundary cell data type below the unit, the upper left corner 156, unit 178 is a data type 1, unit 180 is a data type 2, unit 182 is a data type 3, and unit 184 is a data type 4, and unit 186 is a data type 5, unit 188 is data type 6 Hes, and unit 190 is a data type 7.The purposes that the setting boundary element has known types has two-layer: at first, it can be used as a scale to determine at the interval between the data cell on level or the vertical direction; The second, it can provide automatic calibration function during decoding.When reading, the image file that is scanned of data block contains the two-dimensional array of pixel value.For the value of distinguishing a data unit and next data cell must know in the horizontal direction with vertical direction on the number of pixels of each data cell.Because the data type of each boundary element is known, obtain the information in horizontal vertical space by the number of pixels of checking each boundary element.Automatically calibration is the critical function that boundary element provides.Reflection colored and printed symbols changes, and depends on the printer of use and the time limit that uses.The symbol of the printing of using under long-time situation is subjected to sunlight or the influence of other adverse condition causes that the colour of little point fades.When decoding have the specified data type data cell be scanned the image pixel value time, regulate threshold value according to the situation of fading of colour and operating position.The pixel value that has a known types boundary element by measurement can be finished and regulate threshold value and is used for decoded data units.This is because of the data cell of colour should follow to(for) the colour of special symbol boundary element, if colour has faded away after using a period of time.Above-mentioned disclosed processing has reduced and makes mistakes and data decode method very reliably is provided.
For storing a large amount of text datas, or the flow diagram that the color image data of a compression is created the colored little point symbol of two dimension is shown in Fig. 7.The text 192 of alpha-numerical is converted to a binary data file 200 by 194 step of ascii data format conversion.Similar, the JPEG compression by the color image data file 196 of continuous tone in RGB or CYMK color space also can produce binary data file 200.Binary data file was encoded in 202 steps with color halftoning data cell according to coding schedule of the present invention with each data cell of representing 3 bit binary data.Utilize the Reed-Solomon sign indicating number to use the data cell deconv with coding polynomial expression generation level and vertical mistake correction unit in 204 steps.Have the unit, angle of predetermined data types and boundary element also by around the insertion symbol in 206 steps.The whole symbol that comprises blocks of data elements and boundary element is converted to a Postscript file 208 in 208 steps, it contains the colour and the position of each data cell of describing with the printer interface higher level lanquage of industrial standard, the information of physical size.The Postscript file is sent to 600 high-resolution laser printers 210 of per inch so that carry out the printing of two-dimentional colour symbolism 212.
When reading, with 600 colored ccd scanner 216 scannings of high-resolution per inch colored little point of a two dimension with digital data coding in 214.The result is red with representing for that specified point, green in 218, and each pixel value of the light reflectance of blue light line is created in R (red), G (green), and the image data file that scans in B (indigo plant) color space.At first, image file 218 must be rotated to suitable direction with further data processing in 220.The symbol that is used to print can relative scanning only be scanned the upper left corner boundary element by seeking data type 0 and finds out with respect to the angle at the reference line boundary element edge of scan image and can finish this step with arbitrarily angled.Next step is to calculate the average scan image pixel value of each data boundary unit to obtain the level and the perpendicular separation information of data cell, and the pixel value of each data type is used to distinguish eight different types of data and adjusts threshold value.According to this information, in 224, the pixel value of data block 224 is changed into data cell and binary data and each unit and represent 3.Create digital byte data by mixing three adjacent data cell.Also carry out parity checking 226 steps for each byte.Position with byte of bad parity is recorded, and it can be used for assisting correction process when identifying wrongly written character joint position.Level and vertical check symbol are used for carrying out Reed-Solomon and make mistakes and detect and make mistakes and proofread and correct 228 steps to produce the data file 230 of proofreading and correct.Produce letter-digital text file 232 by traditional ASCII decoding to binary data file 230.For coloured image, software 234 decompression binary data files 230 are created image file by contracting with traditional jpeg decompression.Xie Ma text data file or color document image 236 can be presented on the computing machine CRT or output to printer or be sent to distant place data center with further information processing by LAN (Local Area Network) or Internet at last.
For the present invention of a colored little point symbol of two dimension of text message storage creation also can use in the foreign language text.Owing to be shown in the primitive of the data block 142 of Fig. 5 b are digital binary word joints, standard output (standard exits) becomes digital byte with the foreign language character code, and it is feasible utilizing the present invention to create colored little point symbol for the information of storing outer Chinese character.For example, it need take 3 bytes for the coding of corresponding 1 byte of Chinese character ASCII character.Therefore, for the foreign language text of encoding, with ASCII character 194 in coding schedule replacement Fig. 7 process flow diagram of corresponding foreign language.And ASCII character 232 is replaced to produce the foreign language text by the coding schedule with the coding foreign language in Fig. 8.
The example of coding/decoding step of the present invention is as follows:
A. data cell is encoded:
For example, word " Hello " is
The ASCII coding is as follows:
H：0100 1000
e：0110 0101
l：0110 1100
l：0110 1100
o：0110 1111
Each byte (8) for a letter-numerical character of expression increases a parity bit as least significant bit (LSB).
H：0100 10000
e：0110 01010
l：0110 11000
l：0110 11000
o：0110 11110
From the coding schedule of Fig. 3, per 3 bit binary data are with a data unit byte code
H：010 010 000
(data cell type 2) (data cell type 2) (data cell type 0)
e：011 001 010
(data cell type 3) (data cell Class1) (data cell type 2)
l：011 011 000
(data cell type 3) (data cell type 3) (data cell type 0)
l：011 011 000
(data cell type 3) (data cell type 3) (data cell type 0)
o：011 011 110
(data cell type 3) (data cell type 3) (data cell type 6)
In order to represent word " Hello ", utilized 15 data unit like this, have data byte (colour) sequence as shown in Figure 9.The coded sequence process flow diagram is shown in Figure 10.
B. the generation of error-detecting and correcting unit
The Reed-Solomon sign indicating number that produces false retrieval survey and correction symbol is available in the prior art.Utilize Reed-Solomon sign indicating number generation level in the data block design to improve the prior art notion by providing with vertical software the present invention who makes mistakes detection and correcting unit.
Make mistakes by software and to detect and the generation of correcting unit.
The example that produces verification unit is as follows.With reference to Figure 11, at first check first row of data block.
Known that 96 data cell lists show some data.Each data cell 3 bit binary data coding.
Suppose that first data cell is type 7 (white cells is represented binary data 111), second data cell is type 2 (binary data 010 is represented in the pink colour unit), and the 3rd data cell is Class1 (violet cell is represented binary data 001).
According to theory of finite fields
GF(2 3)
By P (x)=x 3+ x+1 defines a Galois field.
The first data cell type 7 of expression binary data 111 also can be by α 5Expression.
The second data cell type 2 of expression binary data 010 also can be used α 1Expression.
The 3rd data cell Class1 of expression binary data 001 also can be used α 0Expression.
96 given data unit of first row also can be with a polynomial expression P=α 5x 95+ α 1x 94+ α 0x 93+ α nx 0Express.
Utilize the Reed-Solomon sign indicating number, the coding polynomial expression=
(x+1)(x+α)(x+α 2)(x+α 3)(x+α 4)＝x 6+α 94x 5+α 10x 4+α 136x 3+α 15x 2+α 104x+α 15
Figure A9812070400191
The first line data polynomial expression deconv that is encoded
Remainder=r 5x 5+ r 4x 4+ r 3x 3+ r 2x 2+ r 1x 1+ r 0
r 5, r 4, r 3, r 1, r 1, r 0It is the coefficient of remainder.
Suppose r 5=α 3, r 4=α 6, r 3=α 1, r 2=α 0, r 1=0, r 0=α 5
The first row verification unit is a remainder, as shown in figure 12.
Make mistakes and detect and the generation process flow diagram of correcting unit is shown in Figure 13.In fact, inspection unit in school is the remainder of data cell deconv with coding polynomial expression gained.For every row and multiple this process of every column weight.
C. the generation of boundary element
Point out that as above-mentioned data block 102 row * 102 row are at first surrounded by a white data cell (data type 7) ring.
By distributing the established data type to produce the boundary element ring for the steady arm of determining (fixed locators).For example, specifying upper left corner boundary element is and then data type 1 of data type 0, data type 2, data type 3, data type 4, data type 5, data type 6, data type 7 etc.
Whole two-dimentional colour symbolisms are converted into a raster image data file, and vertical scanning direction is the short scan direction, and horizontal scan direction is a slow scan direction.Color laser printer is once printed a scan line image and is moved on to next scan line then until having scanned entire paper on fast scan direction.
Corresponding every kind of chrominance section: cyan, magenta, yellow, black must produce four image files.With reference to Fig. 3, pixel value and data type table, and suppose a single vertical scan line sequentially scan-data type 0 and data type 2 data cells, output pixel value will be as follows:
C.ascii image data file (cyan)
255,255,255,0,0,0 for scan line.
Y.ascii image data file (yellow)
255，255，255，0，0，0
M.ascii image data file (magenta)
255，255，255，123，123，123
K.ascii image data file (black)
255，255，255，0，0，0
A boundary element product process is illustrated in Figure 14.
This identical technology also can be applied to the color laser printer that improves the photographic resolution technology that has of recent exploitation, and this printer contains the pinkish red and blue or green special photograph printing ink than light shading, and comparing with four kinds of colored inks of tradition provides six kinds of colours.Utilize light greenish blue and light magenta ink, and 4 bit binary data coded data unit with enough noise resisting abilities are feasible.
Should notice that any PC with Pentium processor can be used for the software of coding/decoding step with enough speed operations.High-performance workstation, for example Sun Sparc 5 workstations can move this software at a relatively high speed.
The software that is preferably used in the operation computing machine is:
1. the encoding software of writing with C Plus Plus.
2. the decoding software of writing with C Plus Plus.
3. be used for the standard Window programming of user interface.
4. the standard software package of driving laser printer.
5. scan the only standard software package of interface with CCD, for example a large amount of general AdobePhotoshop.
With reference to Fig. 7, piece 192,194 ... the 208th, the software of writing with C++; Color laser printer 210 can be HP color laser ink-jet printer or Xerox Docucolor40 printer.
With reference to Fig. 8, color scanner 216 can be only (600SPi) and a piece 218,220 of UMAX chromoscan ... the 236th, the software of writing with C Plus Plus.
Described the present invention with reference to preferred embodiment, should understand those skilled in the art can not break away from the modification of the spirit and scope of the invention and substitute it.In addition, can under the situation that does not break away from essence of the present invention and spirit, the present invention be made a lot of modifications to be adapted to concrete situation and material.

Claims (12)

1. coded data method in two-dimentional colored little point symbol mode of creating the storage numerical data, this method may further comprise the steps:
Transform letter-digital text and binary data file and become binary data file;
The described binary data file of encoding;
Generation is based on the level and the vertical correcting unit of makeing mistakes of described coding binary unit;
Produce boundary element in response to the described correcting unit of makeing mistakes;
Generation contains the message file relevant for described data unit information;
Be coupled described message file to printer.
2. the method for the two-dimentional colour symbolism of the numerical data of the expression storage of decoding, this symbol comprises border and data cell, and this method may further comprise the steps:
Scan described symbol to produce the image data file of scanning;
Read described border and data cell;
Check image file wrong of described scanning and proofread and correct described mistake;
Produce binary data file; With
From described binary data file, produce coloured image and text.
3. a machine-readable data structure comprises:
Be used for a large amount of numerical datas and store a two-dimensional blocks of data of the color data unit of printing with a plurality of row and a plurality of row, each described color data unit long number binary data coding, a digital byte is represented in the adjacent color data of a plurality of described levels unit, each described row contains data cell and verification unit in the data block, each described row contains data cell and verification unit in the data block, and the boundary element ring surrounds described two-dimensional blocks of data.
4. according to the data structure of claim 3, wherein said structure is formed on the non-volatile read-only device.
5. according to the data structure that proposes in claim 3, wherein each described color data unit comprises by for cyan, yellow, and magenta and black partly have the little point of two-dimentional color halftoning of the same color that the printer of particular pixels value produces.
6. according to the data structure of claim 3, wherein each described color data unit is represented the multidigit binary data and is selected the colour of expression multidigit binary data to satisfy red, the demand of maximum noise resisting ability when reading in the green and blue color space.
7. according to the data structure of claim 6, wherein data are decoded with the reading device of the reflectance of measurement data unit, and this reading device utilizes the light of various colours and the pixel value in the color space and a plurality of threshold ratio are become the multidigit binary data than the signal that is reflected with decoding.
8. according to the data structure of claim 3, wherein said digital byte comprises that a parity bit is used for each byte is detected the mistake of makeing mistakes in the odd number position when reading.
9. according to the data structure of claim 3, wherein each described row comprises data cell and inspection unit, level school, and each row also contains data cell and vertical check unit and is used for makeing mistakes when reading and detects and error correction.
10. according to the data structure of claim 3, the designated unique data that is different from other angle of described boundary element ring in the angular data unit of the selection of wherein said boundary element ring is used for when reading the direct sum image being rotated.
11. according to the data structure of claim 3, wherein the data of each data cell are specified in advance so that provide level and perpendicular separation information during the decoding of the video data block that scans in described boundary element.
12. according to the data structure of claim 3, the known position in the designated described boundary element ring of different pieces of information of the long number data that wherein are used to encode is used to set the pixel value of each data type of threshold values to be provided as decoded data block.
CN 98120704 1998-09-23 1998-09-23 Color semi-color tone micropoint coding method for high density information storage Pending CN1248752A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 98120704 CN1248752A (en) 1998-09-23 1998-09-23 Color semi-color tone micropoint coding method for high density information storage

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 98120704 CN1248752A (en) 1998-09-23 1998-09-23 Color semi-color tone micropoint coding method for high density information storage

Publications (1)

Publication Number Publication Date
CN1248752A true CN1248752A (en) 2000-03-29

Family

ID=5226860

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 98120704 Pending CN1248752A (en) 1998-09-23 1998-09-23 Color semi-color tone micropoint coding method for high density information storage

Country Status (1)

Country Link
CN (1) CN1248752A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1898676B (en) * 2003-12-25 2010-12-01 吉田健治 Information input and output method using the point pattern
CN102142077A (en) * 2010-02-03 2011-08-03 索尼公司 Data reader, data reading method and data recorder
CN101149813B (en) * 2007-07-13 2012-02-29 中兴通讯股份有限公司 Matrix type colorful two-dimensional bar code encoding and decoding method

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1898676B (en) * 2003-12-25 2010-12-01 吉田健治 Information input and output method using the point pattern
CN101149813B (en) * 2007-07-13 2012-02-29 中兴通讯股份有限公司 Matrix type colorful two-dimensional bar code encoding and decoding method
CN102142077A (en) * 2010-02-03 2011-08-03 索尼公司 Data reader, data reading method and data recorder

Similar Documents

Publication Publication Date Title
US5818032A (en) Encoded color halftone micro-dots for high density digital information storage
CN1211754C (en) Method and device for producing and decoding clear visual barcodes
CN1147125C (en) Image coding method, image coding device, image decoding device and forming device
US7478746B2 (en) Two-dimensional color barcode and method of generating and decoding the same
CN1164120C (en) Image encoder and image decoder
CN1617567A (en) Method and device for compressing color images
CN1044348A (en) The method and apparatus of recording halftone video and consequent half tone video
CN1708769A (en) Two-dimensional code having superior decoding property which is possible to control the level of error correcting codes, and method for encoding and decoding the same
CN1215430C (en) Method and apparatus for fault tolerant data storage on photographs
WO2011108605A1 (en) Method for creating and method for decoding two-dimensional colour code
CN1258058A (en) Image processing method and device, image processing system,and memory medium
JPWO2009144794A1 (en) Information code
CN1358028A (en) Image codec method, image coder and image decoder
JP3923796B2 (en) Two-dimensional code, two-dimensional code creation method, two-dimensional code decoding method, apparatus for executing the method, program for executing the method on a computer, and computer-readable recording medium recording the program
CN1828736A (en) Recording and reproducing device
US8614828B2 (en) Print apparatus, server, method for improving print characteristics, and computer program therefor
CN1248752A (en) Color semi-color tone micropoint coding method for high density information storage
US7971789B2 (en) Document with indexes
JP5310502B2 (en) Two-dimensional code, two-dimensional code reader, and two-dimensional code generator
JP4588916B2 (en) Two-dimensional code reading device, two-dimensional code reading method, two-dimensional code reading program, and computer-readable recording medium
JP4645701B2 (en) Image reading apparatus and image reading method
US8958130B2 (en) Method and apparatus for archiving an image
CN1157704C (en) Addressing method for plasma display pannel based on separate even-numbered and add-numbered line addressing
CN1604614A (en) Image processing apparatus
JP2006330930A (en) Method and device for preparing and reading color code

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C02 Deemed withdrawal of patent application after publication (patent law 2001)
WD01 Invention patent application deemed withdrawn after publication