Printing Black and White Reproducible Color Documents

The present invention relates to a method and apparatus for printing color documents in such a manner that when the color documents are reproduced on a black and white copier or printer, their color information content is retained.

A portion of the disclosure of this patent document contains material which is subject to copyright protection, The copy right owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the Patent and Trademark Office patent file or record, but otherwise reserves all copyright rights whatsoever.

Color printers and copiers usually operate in work environments that still include black and white devices. While color printing has become increasingly common in the workplace, the cost of color printing still remains substantially above that of black and white printing. Thus, while color documents can be readily created for small volume use, large volume reproduction suggests that black and white printers or copiers should be used.

When a color document is copied to black and white, the information conveyed by the color can be lost. Information is often conveyed by color through the use of distinct colors representing distinct information. When color documents are copied on a black and white copier, or faxed on a black and white facsimile machine, or reproduced from a color file to a black and white print, information may be lost. These devices usually threshold the luminance component of the source image to represent the image in either black or white, raising the problem of translating the intermediate luminance of colored objects on a page. Many darker colors often copy as black, while lighter colors drop out completely. When gray tones are reproduced, they are often mottled or ragged. Ironically, increases in printer resolution and halftone frequency that improve color image quality act to exacerbate the copy problem. Increases in printer resolution make the image appear to the copier as well as to the human observer as smooth areas of intermediate luminance rather than a pattern of halftone dots.

Even if a copier could reproduce the luminance level of source colors reliably, many different colors have the same luminance and are therefore indistinguishable after copying. This may not be a severe problem for pictorial images where much of the information is in the luminance component, but for text or presentation graphics, color offers new dimensions for conveying information. Color can be used to organize important items, distinguish categories, and organize documents and document content. However, when copied or faxed to black and white, the distinguishing capability of color is lost.

The primary goals of most documents is to convey information. If one can anticipate that the document will be copied or faxed with a black-and-white reproduction device, and information might, therefore, be lost, it might be desirable to sacrifice some of the appearance of the original document for the sake of information content. One might, for example, decide not to use color at all. Alternatively, a selection may be made to use color and color features that can be copied well. Such choices are significant barriers in document creation.

It is known that selected background and/or alphanumeric text areas may be printed in repeating and/or varying patterns of highlight color and interposed with black repeating or varying patterns as disclosed, for example, for single pass xerographic printing of composite image security documents in U.S. Pat. No. 5,291,243 to Heckman et al. Some of such color patterns might also reproduce when such documents are copied on a black and white copier, however, such copying is not reliable.

Of further background interest are U.S. Pat. No. 5,308,553 to Roetling and U.S. Pat. No. 5,153,576 to Harrington. These patents, however, electronically convert a full color document into a monochrome facsimile for subsequent monochrome printing by electronically converting pixel by pixel the original solid colors into various different patterns of black only images. Such an arrangement, by definition, does not produce originals, which are reproduced in color, for subsequent copying on a black and white copier which retains the informational content of the original color image. Further, the patterns that these methods produce are limited either to a fixed set, or to designs directly controlled by only the three color components and where areas controlled by the different color components do not overlap.

Other patents of interest include U.S. Pat. No. 4,703,318 to Haggerty, which illustrates the idea of forming a character based monochromatic image from a digital representation of a color image by: (1) forming a character based representation of the color image in which a plurality of color characters each describe the background or foreground color within a defined area of the color image; (2) providing a plurality of color pattern masks each corresponding to a color in the color image comprising a pattern of light and dark dots capable of providing in the monochromatic image visual discrimination between areas of different color; (3) transforming the color characters into monochromatic characters by replacing the background and foreground colors of each color character in the corresponding pattern of light and dark dots; and (4) displaying the monochromatic characters on a character based output device. The present invention differs from Haggerty in that the goal is to produce a printed image in which is copyable, and in which colors and shades of colors can be reproduced to maintain the informational content of the original document which used those colors.

U.S. Pat. No. 3,807,852 to Hoydic illustrates an arrangement in which color is used to identify particular identified areas for copying on a black and white copier.

U.S. Pat. No. 5,321,470 to Hasuo et al relates to an arrangement which uses color and the processes of a color copier to prevent forgery of color documents by improper copying.

U.S. Pat. No. 5,138,465 to Ng et al relates to an apparatus and method for producing selectively edited reproductions of the original document in which alphanumeric character codes written using highlighted color arrangements designated treatment to be applied to a particular document.

U.S. Pat. No. 5,153,739 to Laman et al the use of different fonts corresponding to different colors. In such an arrangement, the difference in the font itself provides distinction between different colors which copied in black and white.

U.S. Pat. No. 5,098,817 to Voorhees teaches highlighting of text in a manner which will reproduce when electronically copied. This is accomplished by applying a tone to only that area of the document to be highlighted and so that the tone remains in place during copying. The tone is sufficiently dark so as to highlight text but light enough so that the text is readable.

U.S. Pat. No. 5,142,337 to Karidis et al teaches an arrangement for reproducing gray levels in place of colors in a two color printer.

U.S. Pat. No. 5,568,248 teaches that a pattern can be substituted for a color in a document reproduction.

The above identified references are incorporated herein by reference for their teachings.

The present invention is directed to a method and apparatus for producing a varying pattern in colors which allows the use of color in the creation of a color original, and will retain distinguishing color content when the color original is copied in black and white.

In accordance with one aspect of the invention, a color image is reproduced with a set of colors wherein each color is defined in terms of a pattern block. Pattern blocks include a pattern using the color to be reproduced and white in density-related combinations so that the luminance difference between the color and white can be retained when copied or printed with a black printer. The patterns add among colors thus allowing combination colors to be represented in terms of added patterns of colors represented in black as distinct patterns.

In accordance with another aspect of the invention, there is provided a method of reproducing color images that retains color intent including: a) receiving an image defined in terms of color primaries; b) projecting the color image onto a base pattern set, representing the component pattern that will be used, perhaps R, G, B, C, M, Y, W (white) and K (black); c) generating the component patterns for each of the colors in the basis set except white are generated, based on intensity of the color; d) redefining basis color patterns in terms of printing color patterns; and e), combining printing color patterns.

In accordance with yet another aspect of the invention, the basis color patterns may be redefined in terms of printing solely in black.

These and other aspects of the invention will become apparent from the following description used to illustrate a preferred embodiment of the invention and read in conjunction with the accompanying drawings in which:

FIG. 1 describes the overall process of the invention;

FIGS. 2A and 2B illustrate the basic principle of the invention, that different densities can be modeled by certain repeatable patterns;

FIGS. 3A and 3B are examples of a different pattern particularly useful for strong colors;

FIGS. 4A, 4B and 4C illustrate patterns for each of the primary colors, R, G and B

FIGS. 5A and 5B illustrate patterns for black (sometimes, K);

FIGS. 6A, 6B and 6C illustrate patterns for the colorant or secondary colors, C, M and Y.

FIGS. 7A and 7B illustrate how increasingly strong yellow colors are treated;

FIG. 8 shows a functional block diagram of a way that the invention may be implemented, using look up tables for the patterns'

FIG. 9 shows a functional block diagram of an implementation of the invention, where the patterns are generated by screening or dithering; and

FIG. 10 shows a variation in pattern usage when the patterns are printed directly to black.

Initially, the principles of the invention will be discussed, before describing an embodiment of the invention. In accordance with the invention, a method to allow black and white friendly copying, is to impose a pattern in the colors in solid areas of an image, where the pattern has a distinctive appearance in the luminance component of the image. The idea is that upon copying, the pattern remains even though the color is lost. Thus one can tell that the area had color and even distinguish between areas of different colors since they will have different texture patterns. The motivation is similar to that of the above cited Roetling and Harrington patents, which map colors to black-and-white textures that can then be copied. The method described here differs, however, in that it produces colored textures that do not vary greatly from the original color for use in the original color document. Thus the document retains its colored look if it is not copied.

In considering the overall process, reference is made to FIG. 1. Initially an image is received, typically defined in terms of Red, Green, Blue (R, G, B) or Cyan, Magenta, Yellow, (C, M, Y). At block 2, the color image is projected onto a base pattern set, representing the component pattern that will be used, perhaps R, G, B, C, M, Y, W (white) and K (black) Next, at step 4, the component patterns for each of the colors in the basis set except white are generated, based on intensity of the color. At step 6, wherever a color is defined in terms of a non-printing color, that color must be redefined in terms of a printing color. Finally, at step 8, all of the defined colors are combined, so that the patterns for a given printing color represent the colorant required. Initially considering block 2, to generate a pattern while still preserving the color, the method projects the original color onto a basis color set. The basis colors are used in the component patterns. The basis color set might, for example, be the eight colors cyan, magenta, yellow, red, green, blue, black and white, and projection onto this basis will be described, but other colors could be used.

Now referring to the block 2 of the process, initially the white component of the overall color combination is separated out. Thus, instead of printing pink. a pattern of red and white is generated at block 4. For a color expressed in additive red-green-blue coordinates (r, g, b), the amount of white present is the minimum of the coordinate values

The remaining color with the white component removed is

When copied the red (for example) will become black, the white will remain white, and a pattern will be reproduced. The separation is carried out in block 4 in a pattern cell, which is then replicated to cover the colored area. The pattern cell can be any shape that can be tiled, but a rectangular shape is simplest. One could, for example, use a square for the pattern cell and draw the red component of the color in the shape of a vertical stroke, with the width of the stroke reflecting the amount of red present in the color. Light pink would have a thin red stroke while a stronger pink would have a wider stroke, as shown in FIGS. 2A and 2B. To faithfully reproduce a full strength red, the entire (or nearly the entire) pattern cell is filled.

If this is done, it will create a uniform area with no distinguishing pattern; it will just look black when copied on a black and white copier. To preserve some texture, even for full reds, you can either limit the extent of the stroke width to not quite consume the entire cell, as in FIG. 3A or add a white texture within the red stroke when it becomes thick, as shown in FIG. 3B. By placing the white line at the position where the pattern comes together to fill the cell the intensity jump that would occur if the line is introduced suddenly is avoided.

Patterns of vertical and horizontal stripes can be used, and were used in initial implementations of the method, but were found to be too obvious, and tended to interfere with and detract from the image. It was found that a more pleasing pattern is formed by an S shaped curve, shown in FIG. 4A. This curve is joined from cell to cell to make vertical waves rather than stripes. Other pattern shapes used are the cross for green in FIGS. 4B, and diagonal stripes for blue, as shown in FIG. 4C. It will be appreciated however, that many different patterns are possible, each having its own benefits.

The idea of separating the white from the color and using both to form a pattern works well for light unsaturated colors, but dark colors of a certain hues may have little white and differ from one another by the amount of black present. To make black-and-white copies of these colors look different you need to change the pattern with the amount of black present. The black amount is given by:

This differentiation of dark colors can be made by giving black its own shape that grows or shrinks with the amount of black present. Thus the pattern will be composed of a combination of shapes including the shape for the primary color (red, green or blue) and the shape for black. A possible shape for the black content could fill the corners of the pattern cell with black triangles, as shown in FIG. 5A. This gives a diamond shape when the cell is replicated. A reason for choosing this shape is that there is little overlap with the patterns for other colors until the amounts of color or black become large. However, this pattern was found to appear too coarse for grays. Better results were obtained using a finer pattern produced by the addition of a diamond in the center, as shown in FIG. 5B. Other patterns may be possible.

Removing the white component from a color reduces at least one of the r, g, b coordinates to zero, so one might consider making the pattern from the shapes for the other two components and black. However, with this approach the shapes will interfere with one another for many of the colors. The reason is that both shapes can fill out most or even all of the pattern cell and are therefore bound to overlap. The worst cases are the secondary colors (cyan, magenta, and yellow) that require both of the component primaries to completely fill the cell, giving total overlap. We can use smaller shapes and reduce their interference if we use shapes for a primary and a secondary color instead of for two primaries. For example, orange can be made from red (r') and a smaller amount of green (g') but we could combine the green with an equal amount of red to make yellow. The orange would then be formed from the yellow (y=g') and the remaining red (r"=r'-g'). The overall area of the pattern cell that must be colored becomes r'=r"+y rather than r'+g'. This makes it easier to fit in all of the shapes. The strategy then is to fill a pattern cell with a shape for the primary color, one for the secondary color, and one for black, with the remainder of the cell white. This requires that shapes be defined for the secondary colors. One approach is to borrow the shapes from the primaries that form the secondaries. However, it was found that better color differentiation could be obtained by using distinct and different patterns for the secondary colors. The criterion in pattern selection was the minimization of overlap between primary and o secondary patterns that can coexist. A pattern set that works well is the horizontal wave for cyan, diagonal stripes for magenta, and a circle for yellow, as shown respectively in FIGS. 6A, 6B and 6C.

As with the primaries, the secondary shapes can expand to fill the cell. To preserve the pattern when this happens, a thin white path is drawn in the center of the shape for cyan and magenta, similar to that illustrated previously at FIG. 3A. Yellow is too light for a white line to be visible, so in this case a black line can be used. If a solid black line makes the resulting yellow too dark, a dashed black line can be used, or a line composed of red and green dashes, as shown in FIGS. 7A and 8B. Since yellow usually copies as white, some method is needed to distinguish light yellow from strong yellow. One way is to increase the length of the center dashes with increasing strength of the yellow.

A summary of the calculation of the amounts of primary and secondary colors follows. If colors are originally expressed as additive red, green, and blue (R G B) ranging between 0and 1, then the calculations are:

where the values r, g, b, c, m, y, k give how much area of the pattern cell should be given the respective color.

If the colors are originally expressed in the subtractive cyan, magenta, yellow and black (C M Y K) ranging between 0and 1, then the pattern color amounts are calculated as follows:

Note that the pattern cell can contain up to three patterns for any given color, the primary color, the secondary color, and black. Because these patterns share the same cell there will be some overlap; and the color on top will replace the color beneath it. The actual inks for the hidden pattern to will be less than desired and a color shift will occur. For example, if one is producing orange, part of the red will be in conflict with the yellow and there will be a color shift from the desired state where both are shown. The simplest strategy for handling this color shift is to include compensating adjustments in the colors requested. These compensations can be included in the color correction process, which should occur prior to the halftoning step.

If, however, for some reason there is no color correction, or it cannot include the adjustment needed, the adjustments can be explicitly carried out on the color values. The solution is to extend the width of the primary shape (the red wave) to compensate for the overlapping area and then print the secondary color (yellow) wherever them is a conflict. The same can be done for black. When either primary or secondary overlap the black, the black can be extended and the primary or secondary color can be used in the overlap region. The adjustment need only take place on those colors that may be "covered" by other colors.

If, for example, the model is that black is placed first, followed by a color from the set (cyan, magenta, or yellow) and last a color from the set (red, green, or blue), then the second color must be adjusted for the last, and lack must be adjusted for both. The adjustments have the form: ##EQU1##

The functions A.sub.ij give the incremental adjustment to the second color i due to the presence of the last color j. The functions A.sub.kij give the incremental adjustment of black due to the presence of both colors i and j.

If the patterns overlap in some easily modeled fashion, then the adjustment terms may be calculated on-the-fly. Otherwise, the adjustments can be precalculated and stored in a table. Small tables and interpolation may be used to save storage space. Optimal corrections depend not only on the design of the pattern, but also the marking characteristics of the printing device. Even with rather rough corrections, fairly good color correspondence can be obtained between a wide set of original colors and their expression as patterns.

The first step in the process is the calculation of the weights or amounts of each of the plurality of colors in the basis color set as described above. These amounts may be adjusted for anticipated overlap of the patterns. The second stage is the actual generation of the component patterns.

This method of generating colored patterns is algorithmic; that is, there is an algorithm that generates a pattern for any color. The method is not limited to some small set of patterns, or a particular palette of colors. Also, similar colors have similar patterns. This means that sweeps of colors become smoothly changing textures without abrupt changes in the patterns

In selecting a pattern size the pattern cell must be large enough to see, yet small enough to fill small objects and to be inoffensive. For copiers, a cell size of 1/24 inch seems to work well. For faxing, however, the reproduction quality is poorer and a larger pattern cell of at least 1/16 inch is needed.

In implementing the above principles, one approach is to make use of the pattern capability of Adobe PostScript Level II. (A product of the Adobe Corporation, California) The operators for filling graphical objects were redefined such that they first extract the current color, construct a corresponding pattern and install the pattern for rendering, and finally fill the graphic shape. A similar approach might be taken at the Microsoft Windows Graphics Device Interface (GDI), replacing colors with patterns by using the bitmap and brush operations. An advantage of this PDL approach is that it is fairly simple to implement. One need only insert a header that redefines the operators at the beginning of the PostScript document. The disadvantage of this approach is that it may be slow for some images. An example of this implementation is illustrated at APPENDIX Section 1.

A more efficient implementation is based on halftoning and also has a simple implementation, but requires modification of rendering software. The main idea, as explained below, is to not only halftone the ink colors (cyan, magenta, yellow, and black) but all of the basis colors (including the secondary colors red, green, and blue). The halftones construct the pattern shapes for these colors. The inking bitmaps are then constructed by combining the appropriate primary and secondary patterns with logical operations in the third and forth steps. In the PostScript implementation, this process, which includes the separation of the pattern colors into the colorant set and the combination of those contributions for each particular colorant, is implicitly carried out by the PostScript pattern machinery. But for the halftone approach it must be carried out explicitly.

The halftoning is a conventional process where the color values (r, g, b, c, m, y, or k) are compared to a halftone threshold value for that position in the image. For example where r(x, y) is the red color value at position (x, y), T.sub.r is the threshold array for red, mod is the modulus operator, m, n are the dimensions of the pattern cell, and .beta..sub.r is the resulting bit value for the red pattern then the halftoning process is described by:

Similar tests are made for all the other colors.

The threshold arrays are designed so that as the amount of ink increases, the basic pattern is drawn and then thickened. For example, if the pattern cell size were 4 by 4, color values varied between 0 and 1, and the pattern were a vertical stripe, then the threshold array might look as shown:

The next steps, represented at blocks 6 and 8 in FIG. 1, is to combine the patterns. In block 6 the color patterns are separated into the corresponding contributions for the actual colorants. For example, the pattern for the red basis color will actually be produced by printing yellow and magenta colorants. Similarly, green is produced from yellow and cyan colorants and blue is produced from cyan and magenta colorants. Thus the red pattern is actually used twice, once for the yellow separation and once for the magenta.

In accordance with the process implemented at block 8, once we have the colorant (cyan, magenta, yellow and black) separations for all of the basis color set patterns, they can be combined to produce the actual colorant controls. For example, the cyan specifications form the cyan, blue and green patterns must all be combined by logical operations in order to produce the overall control pattern for the cyan colorant. This combination operation is described below. The V symbol indicates the logical disjunction of the bit patterns. ##EQU2##

It may be desirable to explicitly remove the areas where colors overlap with black. You can choose for either the black or color to come out on top, and exlicitly controlling the outcome also avoids the excess ink that would occur if both were printed. The and indicate conjunction and negation respectively. Removing the overlapping color from black would be done as follows:

The bitmaps .beta..sub.k, .beta..sub.k, .beta..sub.m, .beta..sub.y, are then used to control ink marking.

One may wish to include additional white or black areas as part of the pattern, such as white or black lines down the center of the colored areas to preserve the pattern in cases where the color expands to fill the cell. Auxiliary white and black areas can be introduced by further halftoning and logical operations. Suppose, for example, one wished to draw a vertical white line in the center of the pattern when there are large amounts of color i. The white line can be constructed by a halftone cell. The cell threshold would be compared to values for color i and would be set such that the line does not occur until there is sufficient amount of the color for its background, as in the following cell:

If the threshold array for white on color i (i=r, g, b, c, m, y or k) is called T.sub.wi, the halftoning test then generates bit values .beta..sub.wi. The overall white is the combination of all white patterns.

The white bits must be removed from the adjusted primary ink patterns ##EQU3##

Black areas can be generated by halftone patterns in just the same way as white areas. In this case the resulting areas would be given by .beta..sub.ki (i=r, g, b, c, m, y). The overall additional black area would be

This must be added to the black and removed from the other colors. ##EQU4##

Note that black is chosen when black and colored areas overlap, then combine the regular black .beta..sub.k with the auxiliary .beta..sub.k0 prior to removing the black overlap regions from the other colors.

Referring now to the drawings where the showings are for the purpose of describing a preferred embodiment of invention and not for limiting same, a basic functional diagram of the invention of a system for carrying out the present invention is shown in FIG. 8. For one example embodiment, we assume that an image has been generated in colorant space, in terms of cyan, magenta and yellow (CMY space), perhaps by one of the methods described in U.S. Pat. No. 5,528,386 to Rolleston et al., or U.S. Pat. No. 5,483,360 to Rolleston et al., wherein images originally represented in device independent color space by an image creator or imaging device are converted to device dependent or colorant space. The image is originally represented in terms of a plurality of separations, perhaps three or more, and in this case, by a cyan separation, a magenta separation and a yellow separation. For the purposes of discussion, it can be assumed that this image is already color-corrected. Each pixel in each separation, can be either a gray level or multiple level image signal representing an amount of colorant between a minimum or white level and a maximum colorant level.

Initially, upon receiving the signals at a system input, generally indicated as 10, the black component of a color value is removed from the signals. The black component is usually equal to the minimum signal value among corresponding pixels in each separation. Corresponding amounts of colorant, where k=min(c, m, y) are removed, reducing the minimum signal value to zero. Now, the image is defined as the combination of two colors. This step is accomplished at a black removal element 12, which detects the minimum, and then subtracts the detected minimum from each signal. The output of this device will be two colors, a zero signal and a black signal K, equal to the minimum value detected.

At this point, we begin to build multiple overlapping patterns to generate the image. Initially, each pattern is built up in each separation by reproducing pattern repetitions of areas of a size of between of about 1/25".times.1/25", or between 3.times.3 to 20.times.20 pixels, depending on the resolution of the printer at which printing will be desired. Each pattern has a colored portion and a white portion. That ratio of colored portion to white portion gives each pattern repetition its luminance response. However, for many important colors, 100% coverage is required, reducing the amount of luminance differential in the repeated patterns. Thus, instead of reproducing with each of the remaining colorants, shapes representing primaries (r,g,b) and secondaries (c.m,y) are used.

The next step in the process is to find out which of the two remaining colors is larger. Labeled as 14, the two color minimum detector is used to find the minimum of the two remaining colors. The minimum or secondary color value is then converted to an r, g, b value at converter 16. The maximum value less the minimum value is maintained as a c, m, y value. Black is maintained as the black value. Adjustments for pattern overlap can also be carried out at this point if they have not been factored into the color conversion process. Color conversion from c, m, y is a well known process, described, for example, in U.S. Pat. No. 5,528,386 to Rolleston et al., or U.S. Pat. No. 5,483,360 to Rolleston et al.

In the next step, the value of the min, max and black signals are used to determine a pattern, at LUT tables 20a, 20b and 20c. The lookup tables receives an input value representing the gray level C, M, Y value, a value representing the gray level R, G. B and the black value. The LUT outputs a pattern based on stored pattern information from pattern memories 22a, 22b and 22cthat is related to 1) the color, and 2) the density desired. One could also select individual pixels from the pattern by also specifying pixel location modulo the pattern size. In an alternative arrangement, illustrated in FIG. 9, this process can be implemented using a somewhat standard screening process, with screen values loaded to comparators 21a, 21b, and 21c in place of the LUT's, and selected by pixel location, modulo the pattern size, from screen pattern memories 23a, 23b and 23c.

The next step at block 26 is to separate the colors of the patterns according to the colorants actually used in the printing apparatus. Patterns contributing to a colorant are combined for each colorant. The image data is now reproducible in pattern format, however, it is highly desirable to avoid overprinting black and color. Accordingly, black logic 24 is applied to prevent two pixels being generated at the same location by the color pattern signals and the black pattern signals.

In an alternative process, rather than printing the patterns with respective colored inks, the patterns may be printed in black ink directly for copying. In such a case, the image is generated as described above, but each signal is added at the output, to provide a driving signal for a black only printer. Many of the advantages of the invention are retained. It is contemplated that a user interface command may be available to control a color printer to print black only in this scenario. It should be noted that, in such cases, a different set of patterns, which are thinner in nature must be used, so that when a thick, low density signal color area is printed (for example, bright yellow), which would be printed as a solid black area is encountered, the new patterns (as shown perhaps in FIG. 10) provide a lower density level for reproduction.

The disclosed method may be readily implemented in software using object oriented software development environments that provide portable source code that can be used on a variety of computer or workstation hardware platforms. Alternatively, the disclosed image processing system may be implemented partially or fully in hardware using standard logic circuits or specifically on a single chip using VLSI design. Whether software or hardware is used to implement the system varies depending on the speed and efficiency requirements of the system and also the particular function and the particular software or hardware systems and the particular microprocessor or microcomputer systems being utilized. The image processing system, however, can be readily developed by those skilled in the applicable arts without undue experimentation from the functional description provided herein together with a general knowledge of the computer arts.

While this invention has been described in conjunction with a preferred embodiment thereof, it is evident that many alternatives, modifications, and variations will be apparent to those skilled in the art. Accordingly, it is intended to embrace all such alternatives, modifications and variations as fall within the spirit and broad scope of the appended claims. ##SPC1##