CA1082109A - Method for digital color correction in color picture recording systems - Google Patents

Abstract

ABSTRACT

A method is disclosed for the digital color correction of measured color picture values during color picture reproduction. A pattern is trichromatically sensed to create color measuring values which are trans-mitted to an input of a color correction calculator. The color correction calculator is adjusted to obtain desired corrected color values. A succession of signal values which are representative of the color space used during the reproduction process are applied to the color correction calculator to create a set of digitalized color corrected dosing values. This set of digitalized color corrected values is placed on a medium such as punched tape and then applied to a memory in the reproduction system. The corrected color dosing values are then utilized in reproducing a color picture.

Description

108Zi~

SPECIFICATION

~ his invention relates to color picture recording processes and more particularly to recording processes having a digital color correction.
U.S. Patent 2,993,087 illustrates a method for reproduction of color pictures in a device for electronic color correction whereby trichromatic color measurement values obtained from a pattern are subjected to a digital color correction process in order to reproduce correct color separations. The color measurement values of the pattern are replaced by corrected values by use of an electronic memory to allow the desired color dosages to occur during the printing process. Previously, the color measuring values were continuously supplied to a memory -loaded with corrected values and the corrected values were then transmitted from the memory. The functional relation~
ships between the color measuring values and the color dosage values were provided by rigidly wired exchangeable plug cards or terminal strips. The function of these cards as embodied in the wiring was obtained either empirically or theoretically. Therefore, only a course correction could be carried out in many cases since only a limited number of terminal strips or plug cards could be provided and only the finished printed products could be analyzed. -In a reIated copending application o~ Klopsch, Canadian Serial No. 272,146, filed February 21, 1977, a device is disclosed for the production of corrected color 108Z11~9 separations for multi-color printing whereby such a device supplies printed shapes in the form of color separations or multi-color printing while sensing a colored original line-by-line and reproducing line-by-line by use of an interconnected intermediate color connection means.
In the case of this system, a color correction position or location is provided in order to render the entire colored picture visible during the correction process. This position has an electronic color camera to 0 first sense an original, a visual color monitor, an adjust-able color correction calculator connected to the camera to permit the correction of the trichromatic color signals supplied by the camera, and a color converter connected between the color correction calculator and the monitor lS which takes into account the fixed parameters of the respective printing method and adaptation to the colors ~ -of the luminous screen materials of the monitor.
The actual reproduction of the color picture itself in the form of color separations occurs by use of a system coupled with the correction position which consists of a sensor for a second sensing of the original and a record-ing unit for recording the color separations. After the correction adjustment has been completed at the color correction position, the color correction calculator is connected between the sensor and the recording unit.
This system is well suited for the production of color separations since the color correction is rendered .

108Z1~9 visible, only one individual color calculator is used, only one color correction position is needed, and, in the case of a reproduction unit having a second sensor and record-ing unit with an interconnected color calculator, the optimized correction wh'i'ch has once been adjusted is main-tained unchanged permitting a simple and fast correction.
It is an object of the present invention to pro-vide a method for the reproduction of color pictures having the advantages of the'previous suggested system and~ in addition, to facilitate the operation of the system to render the application of such devices to be more flexible and to improve the correction processes.
In accordance with the method of the'invention for color correction which may be'used during color pic-ture pattern reproduction processes, an analogue color measuring value obtained from scanning a color pattexn is digitalized and applied to a memory containing correct-ed color dosage values. These corrected color dosage ~ -' values are then replaced by corrected color value data ' which is fed to a recording unit for recording the color pattern. An analogue color correction calculator is also ~ ' provided. A color picture pattern is trichromatically scanned and the color correction calculator is adjusted.
A succession of digital-to-analogue converted signal values is fed to an input of the analogue color calcula- -' -; tor. These values are representative of a color space used during the reproduction process. The analogue-to-digital converted output signals from the color calculator which result from the input of digital-to-analogue con- -'~
verted signal values are placed into the reproduction ~'' process memory to serve as the corrected color dosage value. - ~

~' 1 '::

: ' . . , , ~ . .:

0821~9 ON THE DRAWINGS

Figure lA is a schematic diagram illustrating a portion of the digital color correction system of this invention wherein a color correction calculator is adjusted and correction data is-generatea;
Figure lB is a schematic diagram of a portion of -the digital color correction system of this invention illustrating a reproduction unit utilizing the correction data from the portion of the system illustrated in Fi~ure lA;
Figure 2 is a perspective view of color values diagrammed to explain an interpolation calculation used with the system of Figures 1 A and lB; and ,.
Figure 3 is a graphic representation of color values in order to explain a two ~imensional interpolat~on ., .

~ ~ ' - 4a - -,. . - - , . . .
.

" . ' ~, . ',, , ~ , ' ~ , .. . .
.

108Zi(~9 ; ~

calculation of this invention.
Figure la illustrates a color correction position having a signal path and principle component units which correspond to the color correction position of the pending Canadian application by Klopsch.
The group of components labeled 1 in Figure la is an arrangement for the rapid line-by-line trichromatic sensing of a color pattern 2 which is mounted as a photo-graphic slide upon a light conductor fiber plate 3 of an electron beam tube 4. The electron beam scans the pattern line-by-line, and the trichromatic primary color measuring value signals R, G, B are obtained from the opto-electric converters 5, 6 and 7 having filters 8, 9 and 10 arranged in front thereof for color separation. These signals correspond to those of the color-camera unit of the Klopsch application. Such signals may be used in place of the ~;
scanning arrangement deseribed above.
It is also possible to sense the original with a -~
so-called drum seanner, a flat-bed sensor, or a density meter.
The analog primary eolor measuring value signals R, G, B are distorted non-linearly, that is logarithmieally, in the three stages 11, 12 and 13. From these stages these analog signals proeeed to an analog eolor-eorreetion caleula- -tor 15 through a switeh 14 and reaeh a eolor monitor 20 via a color eonverter or deeoder 16 and amplifiers 17, 18 and 19. The color ealculator, the eolor converter and the eolor monitor are - . . . ~ -1~8Z109 shown in Figures 3 and 4 in the Klopsch application, and therefore the circuit technology is not shown here. The primary color measuring value signals R, G, B are con-verted into the printing-color signals magenta, cyan, yellow and black (Mg, Cy, Yel, and Bl) which are later required in the reproduction unit described in Figure lb. In order that the colors are indicated on the monitor 20 in the manner in which they will later appear in print, the con-verter 16 converts the color dosage signals Mg, Cy, Yel and Bl back into the color signals R', G', B'. This calculation can be carried out with the help of the linear computation matrix having the following form:
all Cy + al2 ~g ~ al3 Ye + al4Sw a21 Cy + a22 Mg + a~3 Ye + a24Sw a31 Cy + a32 Mg + a33 Ye + a34Sw This matrix may be realiæed by providing an addi-tion resistance matrix or by use of operational amplifiers as shown in the Klopsch application, Figure 4.
For the negative coefficient a, the signals Cy, Mg, Yel and Bl are also made available with a negative sign. The coefficients are adjusted in such a way that the ~-colors of the quadruplet Cy, Mg, Yel and Bl which appear - upon the color monitor 20 will coincide. For this purpose a random color chart may be scanned via the scanner of the color correction position and then further processed for finished printing by use of the reproduction unit shown in Figure lb. If the color chart is then placed next to -i ' ' .
.

the color screen of the monitor 20, and even if an average brightness is obtained upon the screen and upon the color-printing chart such as by way of additional illumination around the color screen, then the coefficients "a" may be controlled and possibly subsequently adjusted for other printing colors. The coefficients thus usually remain in fixed adjustment.
If during the scanning of a desired picture pattern 2, the correction is adjusted by parameter regula-tion of the calculator 15 in such a way that the picture onthe monitor 20 appears with correct coloring, then the switch 14 is switched. An impulse generator 21 is started which causes a counter 22 to begi.n to count, for instance from 0 to 511. These l-dimensional number successions of 29 are converted into a three-dimensional succession of 23, thus 3 times 8 stages, which are processed to a digital-analog converter 24 via the line group 23 and, further, to the color calculator 15 as equally spaced sig-nal values of eight positions, respectively, via switch -14 in all possible but ordered combinations. The output signal Cy, Mg, Yel, and Bl are subsequently digitalized in analog-digital converters 25, 26, 27 and 28 which are connected to the outputs of the color calculator 15 and are processed onto punched strips 30 in a strip puncher 29 and connectéd to the A/D (analog digital) converters.
In place of the arrange~ent described above, it -is also possible to provide an arlangement whereby a -. '.:
'~

- - . . .
, . ~

1~8~ 1~ 9 measuring instrument indicating the color dosage values is utilized in place of the monitor and whereby a television camera replaces a trichromatic sensor. In such an arrangement, a sloweP analog or digital converter may be used instead of the fast color correction calculator.
The digitalizing occurs in binary fashion with the simultaneous input of a command signal via the line 22' from the counter 22 or from the impulse ~enerator 21.
For example, ~ x 12 printing-tone values will be stored at for instance 8 bit = 256 steps. These values may, of course, also be processed into another intermediate memory or into the memory of the reproduction unit described in Figure lb.
The above described process has the advantage that the correction behavior of ~he color correction cal-culator 15 is arrived at by visual control through use of the monitor during a sensin~ of the colored pattern 2 ; by individual adjustment for each color value within the color space. The correction data are stored upon the punched strip which may be used in a reproduction unit for controlling color correction which is entirely independent ~rom the color measuring position.
In place of the punched strip, any other data carrier means may be utilized or the data may be trans-mitted directly to the reproduction unit.
Figure lb shows a reproduction unit which, inthe present example, represents a device for the production ; -1 ~21 Q9 of color separations. The mechanical design is similar to that shown in Figure 1 of Klopsch application at refer-ence numeral 2. The reproduction unit consists of a scanning drum 32 driven by a motor for receiving a pattern

2 and a recording drum upon which the color separations 34 are produced. A raster disc is positioned upon the common shaft 35 and supplies synchronous timing impulses to the timing generator 40 via the light beam formed by devices 37 and 38 and an amplifier 39.
The scanning unit 41 associated with the scanning unit 32 and which has the same spectrum sensitivity as the , scanning unit 1 of Figure la, supplies the three primary , color measuring value signals R, G, B during the rotation , of the scanning drum. These signals are processed to the ~
A/D converters which are controlled by the timing gener- ,' ' ator 40 via logarithmic converters 42, 43 and 44 which , ' correspond to the converters 11, 12 and 13 of Figure la.
There they are converted into digital signals corresponding -' to the color tone values produced in Figure la. Thus, the ~,, 20 uncorrected digital color measuring value signals are ' available at the output of the AtD converters 45, 46 and -47.
, The correction process is as follows. The punched ~ -; strip 30 with the correction data,of the entire color space , 25 is read out with the help of a punched strip reader 49 and processed into the digital correction memory 50. In order to correlate the addresses with the information, for ,':
.
- ~-, .
. . .
~. . .
_9_ - , . - . ...

10821()~
instance the color tone value states with the tone values themselves, a counter 51 moves with the punched strip and outputs a corresponding timing rhythm to the address input of the memory 50 via a switch 52. After the corrected color dosage values are inserted, the switch 52 is acti-vated for carrying out the correction in synchronization with the rotation movement of the disc 36.
The most straight forward method of correction would now be to switch the uncorrected signals from the A/D converters 45, 46 and 47 into the memory 50 and to replace them by corresponding correction data in the memory. In the case of 256 steps of tone values, for which an 8-bit coding is required, a total of 224 tone values will occur. The digital correction memory thus would have to contain 224 memory positions. Although this is techrlically possible, a high storage expense results.
~ For this reason, the means of interpola~ion previously known in mathematics is utilized such that only a support-- ~ ing frame-work of color tone values is stored in the intermediate values obtained by calculation. In the example of Figure lb, the memory 50 contains a supporting framework of color dosage values for the individual color separations. The data supplied by the A/D converters 45, 46 and 47 consists of an 8-bit combination, for example.
The four higher bits are split in a separation stage or switch 53 and switched into the memory 50. The color dosage values which belong to the respectively inserted , , ~08ZlQ9 value will then appear at output 51 at the output of the memory.
In order to obtain the color dosage values which result in the system of points in. the color space defined by the three R, G and B values, an address-increasing unit 55 is provided which can increase the addresses of the three partial addresses at the 4th bit by the amount of "1". Thus, eight different addresses may be applied to the memory and a total of eight values then appear at the out-put 54 of the memory 50. In the three dimensional color space, the corner points of the cube are thus provided to surround the image dot itself which is to be reproduced.
Due to linear interpolation, the color dosage value may be computed in the interpolation stage 56. The three low value bits are supplied to the interpolator 56 via lines ~ ~
57. They represent the distances of the picture value ~ -which is to be reproduced from the corner points of the - :
cube of eight correction values.
Since the interpolation must occur faster than the scanned picture dot succession, a timing multiplier is connected to the timing generator 40 which is a compo-nent of the control mechanism 58 for the interpolation process. The result of the interpolation is processed - -towards a D/A converter 60 which controls the recording unit 61 and causes the correspon~ing color separation 34 ~ :
~o be recorded upon the drum 33.
In another application of the invention mutually - .

,; ,~ -11-,, , . : .. .. . : .

~08~109 deviating color corrections can be carried out in different parts of the picture. With such a system, the first color correction is carried out several times and the color dosage values are stored several times so that the memory 50 is present in multiple form. During the reproduction, a switching operation is carried out between the different memories depending on the different picture parts being reproduced.
The above described swi tching may be carried out with the help of a mask (U.S. Patent 3,629,490), with the help of a color recognition circuit (German Offenlengungs-schrift P 25 44 703), or by way of providing coordinates or fields to permit color recogn ition circuits to be used for switching over between the memories.
A preferable interpolation scheme in the present invention is shown in Figures 2 and 3.
Figure 2 shows a cube resulting from color dosage values which are read out from the memory 50 by the higher value bits and the higher value bits at the address in-crease by "1". These are the following eight values:
FR,G.B~ FR~G"B~ FR'G'B"' FR~G"B"
FR"G"B" ' FR"G"B " FR"G"B" ' FR"G ' B ' The coordinate values R, G, B, which result from the four low-value bits and which are used for interpola-tion, are drawn within the cube forming this value. The interpolation calculation proceeds such that first the color value FR,GB is interpolated from the four color .
, ~; -12-: . , - ~ .. . . . .. .
' ~ , 1(~8~109 values of one of the cube planes, here the lower one with the joint index P~' and the values _ and G. Parallel thereto the interpolated value FR"GB is calculated from the four color values with the index R" and the values _ S and B. The desired interpolation value FR - R FR..GB + (1 - R) FR'GB-is calculated from these two partial interpolation values with the help of the value R.
Figure 3 explains the preceding partial interpo-lations . The four values FR ' G ' B ~, FR ' G ' B" . FR ' G"B ~ and FR,G"B" are shown as vertical amplitude lengths over the corners of a basic square. Depending on their origin, the values G and B assume binary sta~e values between zero and -one. Accordingly, the following results in three interpo-lation steps:
1) x=B FR~G..g-- + (1 - B) FR'G"B' 2) y=B FR,G,B" + (1 - ) FR'G'B'

3) FR.GB = G x + (1 - _) y The following also results:
1 - 3): FR,GB = GB (FR~G~B~ + FR'G"B" R G B
FR.G..B-~ + B (FR~G~B" ~ FR'G'B') + G (FR~G~B~ ~ Fæ'G'B') + FR'G'B'-The second partial interpolation for FR"GB has the same form as the index R".
With the invention, the advantage results that the actual color correction can be performed in advance and stored, and that a better utilization of the ~achines is ' .
,' ,, .

; -,- , - ,, , . ... , :, :, .
. ~ . . ,, .. . . .:
.. . . . : .

possible. Thus, a complete separation of adjustment and processing can occur.
It is also possible to use an untrained worker at the reproducing device who does not have the specific S knowledge needed for color correction.
It is ~urthermore possible to produce and store correction data for later application in the same or simi-lar patterns.

' '.

~ -. .

i ., 2s ..:-:

.~ ....... ..
-. . . .. . . .

Claims (7)

The embodiments of the invention in which an exclusive property or privilege is claimed are defined as follows:

1. A method for color correction which may be used during color picture pattern reproduction processes in which analog color measuring values obtained from scanning a color pattern are digitalized and applied to a memory containing corrected color dosage values, said corrected color dosage values then being replaced by corrected color value data which is fed to a recording unit for re-cording the color pattern, the method comprising the steps of: providing an analog color correction calculator;
trichromatically scanning a color picture pattern and adjusting the color correction calculator; and feeding a succession of digital to analog converted signal values to an input of the analog color calculator which are representative of a color space used during the repro-duction process and placing analog to digital converted output signals from the color calculator which result from this input of digital to analog converted signal values into the reproduction process memory to serve as said corrected color dosage values.

2. The method in accordance with claim 1 in which during the adjusting of the color correction calcu-lator a monitor connected at the output of the color correction calculator indicates a color corrected picture pattern.

3. The method in accordance with claim 1 in which a portion of said signal values representative of the color space are placed into the color correction cal-culator and subsequently into the memory as corrected color dosage values, and that to obtain intermediate values of the stored corrected color dosing values an interpolation is carried out between the stored dosage values.

4. The method in accordance with claim 1 in-cluding utilizing several memories in order to simulta-neously obtain corrected color dosing values for several portions of the color picture pattern.

5. The method in accordance with claim 1 characterized in that in order to carry out mutually deviating color corrections in different parts of a color picture pattern the adjusting of the color correction calculator is carried out several times, and the corrected color dosing values are placed into different memories, and that a switching is carried out between the different memories during picture reproduction depending on the different picture pattern parts being reproduced.

6. The method in accordance with claim 1 characterized in that the color corrected dosing values are intermediately stored upon a data carrier after the adjusting of the color correction calculator and before being placed into said memory.

7. A method for color picture reproduction, comprising the steps of: prior to a reproduction providing an analog color correction calculator and color monitor;

trichromatically scanning a color pattern, feeding the resulting analog primary color measuring value signals through the color correction calculator and monitoring the resulting color pattern signals on the color monitor;
adjusting the color correction calculator for desired color dosages; feeding a succession of digital to analog converted signal values representative of a color space used during the reproduction process through the color calculator and digitalizing the resulting output which serve as corrected color dosage values; storing the corrected color dosage values in a memory; and in the reproduction process: scanning a color pattern to be reproduced to create analog color measuring values;
digitalizing the color measuring values and feeding them to the memory containing the corrected color dosage values to create an output from the memory of corrected color value data; and using the corrected color value data for recording the color pattern intended for re-production.