Method and generator for superposing graphic patterns.

The present invention relates to a method and generator for superposing graphic patterns, each of which is represented by a rectangular array of print elements (pels or dots), for use in presenting documents on a cathode ray tube display, printing device or other image or document presentation apparatus.

Superposing graphics, such as an underscore with an alphabetic or numeric character pattern, has been achieved by so-called overstriking. Overstriking is printing or displaying two character graphic symbols in one character graphic space. For example, a lower case "g" pattern can be underscored by overstriking with a horizontal line underneath the descender portion, through the descender portion or by breaking the horizontal I ine only at the outward facing boundaries of the "g" descender. For typographic quality, it is desired that the underscore be broken at all inward and outward facing boundaries of the descender portion with the break being contoured to the descender portion and separated therefrom in accordance with the weight of the "g" character. While ligatures can be constructed to reach the typographic quality desired, computerized printers and displays, such as cathode ray tube displays, electrophotographic printers and the like, employing rectangular arrays of dot patterns have required manually creating raster patterns for implementing the underscore to be overlaid or superposed over the character graphic. As the terms are used herein for underscore purposes, the character graphic is a dominant or superior graphic whi le the underscore is a recessive or inferior graphic to be interrupted in a manner contoured to, but spaced from, the superior or dominant graphic by an arbitrarily selected distance. Such graphic superposition is fully automatic, such as can be employed in a programmed digital computer.

Superposition of graphics also has been achieved by creating an occlusion signal in video circuits whereby the inferior image is blanked from the video picture while the image being superposed is placed in the occluded area. U. S. Patent 3,961,133 shows such an occlusion signal wherein each line of the raster pattern has an occlusion signal measured in the time domain to make room for the image being superposed on a larger image on the video screen. Such a system does not provide for arbitrary and selective spacings between the images which are contoured to one of the images. Generation of character graphics, including overlapping, is shown in U. S. Patent 3,984,828. This patent uses a bar or rectangular array approach to create a plurality of nume- rics, none of which allow contouring with selective spacing between superposed graphics.

Similarly, U. S. Patent 4,317,114 shows superposing one graphic on another using an inhibit or NOT function wherein ne image is superposed on the other by blanking out the other image, i.e., similar to the first-mentioned patent ₃,₉61,133. Examination of this reference showed no selective spacing of one graphic from another for achieving the typographic quality superposed graphic, as mentioned above. Accordingly, there is needed an automatic means of superpos- i_ng graphic patterns generators which facilitates the generation of typographic quality superposed graphics of all types, particularly for underscoring, no limitation thereto intended.

In accordance with the invention, a superior and inferior graphic are selected to be superposed one on the other with both graphics being respectively represented and generated by rectangular arrays of binary signals, one binary signal representing the graphic (foreground pels) while the other binary signal representing background (background pels). An array of digital signals is generated from the dominant graphic array by adding foreground pel indicating digital signals to the dominant graphic array in accordance with the desired graphic separation. The generated array is preferably enlarged by a separation dimension equal to an integral number of dots or pels in a circular array around each foreground pel in the dominant graphic pattern. The generated arrav then masks the inferior graphic array to create an occluded or fragmented array. The fragmented array is then ORed with the dominant array to create a superposed graphic pattern which is a combination of the two original graphic patterns and having a predetermined separation dimension therebetween, contoured to the dominant graphic pattern. The resultant array is usable to generate a typographic quality superposed graphic.

For superposition efficiency, only a portion of the dominant graphic pattern need be used in the superposing processing. The inferior graphic pattern is measured, then a portion of the dominant graphic pattern equal to the size of the inferior graphic pattern plus the separation dimension is selected for creating the intermediate raster patterns with a final ORing step, which includes the entire dominant graphic pattern. Other variations on selecting array sizes are practiced within the present invention for obtaining a diversity of graphic superposition results.

The foregoing and other objects, features, and advantages of the invention, which is defined in the attached claims, will be apparent from the following more particular description of the preferred embodiments of the invention, as illustrated in the accompanying drawings, of which <ul id="ul0001" list-style="none"><li>Figure I is a block diagram of a graphics preparation system employing the present invention.</li><li>Figure 2 is a simplified flowchart of machine operations performed in the graphics preparation system of Figure 1 used in explaining the operation of the system in practicing the present invention.</li><li>Figure 3 is a diagrammatic showing of expanding a dominant graphic pel or dot for practicing the invention in the Figure 1 illustrated system.</li><li>Figure 4 illustrates a rectangular array expansion from each pel of a dominant graphic for practicing the invention in the Figure I illustrated system.</li><li>Figure 5 shows a sequence of dot or pel patterns resulting from a series of machine steps of the Figure 1 illustrated system during a superposition of a two-pel array on a larger inferior graphic raster pattern.</li><li>Figures 6 and 7 illustrate superposition of two different inferior graphics on a single dominant graphic as superposed by the Figure 1 illustrated system.</li><li>Figure 8 is a chart of machine operations for the Figure 1 illustrated system for practicing the present invention.</li><li>Figure 9 is a diagrammatic showing of machine operations for implementing the Figure 8 illustrated flowchart.</li><li>Figure 10 is a flowchart illustrating some of the machine-implemented steps of the Figure 9 illustrated machine operations.</li></ul>

Referring now more particularly to the drawings, like numerals indicate like parts and structural features in the various diagrams. Figure 1 shows an interactive terminal 10 having a usual CRT display screen 11 and an input keyboard 12 coupled to a programmed digital computer 13 via a suitable coupler 14. Digital computer 13 is shown in a logical form for illustrating how various programs in the digital computer 13 interact with the terminal 10 for implementing the present invention. It is to be understood that the electronic circuits of the digital computer 13 are those circuits usually employed in constructing a programmable digital computer of a general purpose type.

Within the computer 13 is a support program 20 which operates interactive terminal 10. Support program 20 receives input from the terminal 10 for use in the superposition of graphic generator pattern arrays in accordance with the present invention. It is preferred that digital computer 13 be of the multiprogram type such that job entry system (JES) ₂₁ can selectively schedule operations of interactive terminal I 10 with other text and graphic or other programs residing within digital computer 13. For example, text and graphics processing program 22 can be intimately related with the operation of terminal 10 or can process text and graphics of all types from diverse sources. Text and graphics processing program 22 includes document formatting which may call out graphic characters such as those generated in accordance with the present invention. 1/0 and output writers 23 is a program that couples the text and. graphics processing program 22 to store the document in an electronic form on DASD (direct access storage device or disk file) 25 as well I as supplying the document in electronic form to an output document presentation device 26. Device 26 can be a typesetter, printer of any type, electronic display such as display 11 for terminal 10, magnetic tape, another DASD or a communication link (COMMO). Library facility 24 is a prograr_l that accesses electronic fonts (sets of binary signals representing character graphic generator patterns) stored in DASD 25 for use by text and graphic processing program 22. To support all of the above interactive text processing, a graphic facility program 30 enables the creation and insertion of such electronic fonts (font generation patterns) into DASD 25 using JES 21 for scheduling an I/O and output writer program 23. The electronic fonts are usually consist of raster arrays of binary digital signals corresponding to the graphic to be presented by device 26. Within the graphic utility 30 are several programs, such as a graphics smoother 31 which rounds the corners of graphics being created and rotator 32 which rotates the raster pattern such that the graphic can be presented in any orientation by the document presentation device 26. Scaler 33 adjusts the point size (vertical height) pitch and width of a character graphic or other graphic to be presented. Resolution program 34 adjusts the resolution of the graphic pattern to the device 26 resolution; for example, an electronic font consisting of a plurality of character graphic patterns can have its resolution changed from 360 pets per inch to 240 pets per inch or to 600 pets per inch, all under control of the graphic utility.

The present invention is embodied in the graphic generator utility 30 by the superpose program or module 35 which operates in accordance with the machine operations chart shown in Figure 2. The graphic utility has the ability to fetch graphic patterns from DASD 25 and supply them to the superpose program 35 and then to store a new graphic generating pattern created by the superpose program 35 in DASD 25 such that the text and graphic processing program 22 can identify such new graphic pattern for use by device 26. In this manner, the graphic utility automatically creates graphic generation patterns based upon previously created graphic patterns for facilitating document presentation with a wide variety of character and other graphic forms.

The Figure 2 illustrated superposition machine operations chart shows step 40 as first expanding the inferior graphics pattern. This step effectively adds background pel signals to the original inferior graphic pattern. The expanded inferior graphic pattern is used in step 41 to identify the segment of the superior or dominant graphic pattern to be used in creating a partial superposed graphic segmentation of the superior graphic pattern. This segmentation reduces the computer processing power required for superposition, i.e., those portions of the graphic patterns that are not affected by the superposition are not processed. At step 42, the segmented superior graphic pattern is manipulated in accordance with the invention to create a mask or occlusion array in accordance with the segmented superior graphic plus a desired physical separation between the superior and inferior graphics in the graphic pattern being created. The mask is then NOT ANDed with the inferior graphic pattern for creating a fragmented graphic array, which is then combined with the superior graphic in step 43. The term "NOT AND" means that foreground pel indicating signals residing in the same array position as occlusion pel indicating signals are switched to be background pel indicating signals which leaves only foreground pel indicating signals of the inferior graphic pattern that coincide with background pel indicating signals of the occlusion pattern. The combination function of step 43 can be two steps. First, the segmented inferior graphic is combined with the segment of the superior graphic used in processing. The resultant graphic then is inserted into the array of the original superior graphic. It should be remembered that the original graphic patterns are still stored in DASD 25; the newly created graphic pattern is stored in DASD 25 in a data storage space specialty allocated therefor using usual data storage allocation procedures.

The steps of Figure 2 are preferably implemented in the superpose graphic program 35 using an APL language such as described in publication GC26-3847, published by International Business Machines Corporation, Armonk, New York as file number S370-22 (July, 1978) and entitled "APL Language". APL has a capability for handling arrays in an efficient manner. Accordingly, it is a preferred language for implementing the invention.

Figure 3 illustrates a process of creating an occlusion pel pattern mask for a single dominant foreground pel 50; it is to be understood that in an array of pets for generating a character graphic, that each of the foreground pets in the graphic will have a similar circle 51 of occlusion wherein no foreground pel of an inferior or recessive graphic is permitted to occur. Figure 3 illustrates an expanded occlusion array for separation between a dominant and recessive graphics of three pels (TTC=3). The general procedure is to generate occlusion foreground pels vertically, then horizontally for enlarging the array of foreground pets around the single dominant pel 50 such that all pel positions within a circle having radius TTC and centered on each foreground pel will contain occlusion pels. The array of digital signals represent a pel array having pets in predetermined pel positions with a constant spacing. First, as represented by arrows 52, three occlusion pets above and below the single dominant pel 50 are added along the vertical ordinate. Then the process generates the occlusion pels in the left semicircular portion of circle 51. The left double-headed arrow 53 represent the addition of seven occlusion pels to the expanded pel array. One of the pets is on the horizontal ordinate of pel 50, three occlusion pels are above the ordinate and two pels below the ordinate. The next step is to add five occlusion pels as indicated by the left arrow 54; one occlusion pel on the horizontal ordinate and two occlusion pels above and below the ordinate. Finally, at 55 one occlusion pel is added at the horizontal ordinate along the circle 51. The right semicircular portion of circle 51 is generated in a similar manner.

Since it is desired to automatically process all graphics as a rectangular array of pels as opposed to a circular array, the rectangular array of Figure 4 is the electrical representation of the occlusion array of Figure 3. The single foreground dominant pel 50 remains in the center of the rectangular array. The Xs represent background pels (binary zeroes) while the encircled Xs 59 represent the occlusion or masking pels added around single pel 50 for creating a spacing between the dominant and recessive graphic patterns.

Figure 5 is a simplified showing of using the principles set forth in Figures 3 and 4 for superposing an inferior graphic pattern 72 on a superior graphic pattern 70 consisting of two foreground pels along one line. For simplicity, background pels are not shown. Circled "Xs" 69 around the two pets ₇₀ corrcspond to circle 51 of Figure 3. The circled Xs 69 have an effective radius of TTC=3 around pels 7₀. Numeral 71 denotes TTC expanded (with rounding of the pel inclusion calculations to include pels on the circular periphery, i.e., make the intergraphic spacing constant for a given spacing value) superior graphic for creating an occlusion or mask pattern is illustrated. The pattern includes the two foreground pels 70 in the center of the array, surrounded by a plurality of occlusion pets 69 represented by the encircled Xs. The numerals on the right-hand side of the arrays indicate the displacements of the rows from the foreground pel 70 baseline represented by the numeral 0. The inferior graphic 72 is shown as a portion of a horizontal line having a thickness of nine pels. The occlusion pattern 71 is logically NOT ANDed with the inferior graphic pattern 72 to produce the intermediate inferior graphic pattern 73 resulting in deleting foreground inferior pets 74 (represented hy the Xs in this portion of the Figure) from the inferior graphic pattern leaving inferior foreground pets shown as dots 77 in the rows +1,+2, +3, +4, -1,-2, -3, -4 of the inferior graphic pattern. Once the masking is completed, the intermediate inferior and the superior graphic patterns are combined in a logical OR fashion to create a new or resultant graphic pattern 75 which includes the original foreground dominant pets 70 bracketed by the remaining inferior graphic foreground pets 77. The background pels are not shown for purposes of clarity. New graphic pattern 75 has the inferior graphic 77 separated from the superior graphic 70 by the distance of three pel positions in all directions. The separation is contoured to the shape of the foreground pets 70, that is, the area represented by double-headed arrow 76 includes pets deleted in order to achieve the desired pel spacing between the graphics being superposed. Such contouring will become more apparent from Figures 6 and 7.

Figures 6 and 7 show the letter "O" first being underlined in Figure 6 and then having a slash extending therethrough in Figure 7. In Figure 6 the solid dots represent foreground pels background pets are omitted for clarity. A first set of pets 80 represents the letter "O" while pels 81 represent an underline or inferior graphic which is to be superposed on the superior graphic 80. The solid dot pels show a separation of two pels between the superior and inferior graphic patterns. The two pels 82 inside the graphic pattern 80 are a portion of the underlining 81. For completing contouring of the underscore or inferior graphic to the superior graphic zero, the inclusion of pels 82 within the superior graphic for achieving typographic quality of superposed graphics is provided by the present invention. To further illustrate the effect of changing the spacing between the superior and inferior graphic patterns, the pels 83 and 85 represented by Xs for showing a separation of the superior and inferior graphic of a single pel position, In addition to leaving an additional line of pels of the inferior graphic pattern 81 facing and contoured to the superior graphic pattern 80, additional pels 84 at the lower curves of the superior graphic pattern 80 are added to more fully maintain the separation between the superior and inferior graphics constant for achieving typographic quality. Such additional pets 84 occur because of the described circular masking for each of the individual foreground pels of the superior graphic 80 as shown by the addition of the occlusion pets in the superior graphic 80, as explained with respect to Figure 3. Additionally, inside the superior graphic pattern 80 the foreground inferior pels 85 are added for the reduced spacing between the superior and inferior graphics results in maintaining typographical quality contouring of the interrupted inferior graphic 81 to the superior graphic 80. It is to be appreciated that the number of pels in a character graphic to be printed will have a much higher density than that shown in Figure 6, thereby the edges of the graphics 80 and 81 will have a much finer and acceptable typographic quality because of such higher resolution.

When inferior graphic 81 is to be superposed with respect to superior graphic 80, it is not necessary to process the entire superior graphic 80. Accordingly, prior to the steps shown in Figure 5, the inferior graphic 81 is measured as to its vertical extent as indicated by double-headed arrow 87. Assuming a desired spacing between the graphics of two pels, then it should be apparent the pels of the superior graphic 80 that effect deletion of pets of the inferior graphic 81 are those pels spaced from the inferior graphic by the desired spacing. Accordingly, as represented by double-headed arrow 88, that portion of the superior graphic 80 lying within the vertical extent defined by double-headed arrow 88 (the vertical size of the inferior graphic plus the desired separation on both vertical boundaries of the inferior graphic) are used in the masking operation shown in Figure 5. Upon completion of the new graphic as shown at 75 of Figure 5, that portion of the superior graphic 80 plus the inferior graphic 81 are reinserted into the total rectangular array of the superior graphic 80 for completing a new graphic. Segmentation of the superior graphic based upon the extent of the inferior graphic for generating a new graphic reduces processing time and data storage requirements.

Figure 7 illustrates (no background pets shown) a slash line 91 being superposed upon a "O" graphic 90 wherein at lower left-hand corner the "O" graphic 90 is the superior graphic while in the upper right-hand corner the slash line 91 is the superior graphic. Accordingly, the resultant new graphic finds the slash line 91 being interrupted by the "O" graphic 90 and spaced therefrom at 92 by a single background pel (not shown). In a similar manner, in the upper right-hand corner, the "O" graphic 90 is interrupted and separated from the slash line graphic 91 by a single pel space 93. The spaces 92 are contoured to the graphic 90 while the spaces 93 are contoured to the graphic 91.

The array expansion follows the teaching of Figures 3 through 5, the graphics are divided into four quadrants indicated by coordinate lines 95, 96; each quadrant of pels is handled independently. The background pels are omitted for purposes of clarity. Quadrant I, the upper right-hand quadrant, (includes spaces 93) is expanded by one pel position on all four corners of the quadrant subarray. That is, the inferior graphic 90 (for that quadrant) is expanded and includes the pels of the quadrant I inferior graphic 90 from the other three quadrants II, III and IV. The subsequent processing for the first quadrant follows the previously described procedure. For Quadrant II, the lower right-hand quadrant, graphic 90 is the superior graphic with the expansion of the graphic 91 as an inferior graphic following the previously described procedures. Similarly for Quadrant III, the graphic 90 is the superior graphic while graphic 91 is the inferior graphic. For Quadrant IV, the quadrant 90 is again the superior graphic. Accordingly, using the segmentation described for Figure 6, applying it to plural portions of the graphics pattern, any number of segments of the pattern can become a superior graphic or an inferior graphic with each segment being expanded as if it were a separate graphic pattern. Recombination of the segments into a single graphic pattern is merely by logically ORing the individually processed segments.

Figure 8 illustrates a generalized flowchart for the computerized or machine generation of a superposed graphic having contoured separation. At 100, the inferior graphic IG is measured such as shown in Figures 6 and 7. At 101, the measured inferior graphic is expanded by a user inputted desired spacing in pel positions between the superior and inferior superposed graphics, the spacing herein abbreviated as TTC. At 102, that portion of the superior graphic SG, hereinafter referred to as segment "A", having a dimension equal to the dimension of the expanded inferior graphic is selected. At 103, the segment A of the superior graphic expands as shown in Figure 4 with the expansion arrays being stored in computer registers (in main memory and not shown) referred to as TEMPS and TEMP4. TEMP3 and TEMP4 have been set to zero such that all the pets represented by these binary signal arrays are background pels. At 105, the segment A portion of the superior graphic is expanded as shown in Figures 3 and 4 and converted into a mask at 106 by adding the occlusion pels, such as pels 59 of Figure 4 to the pels already existing in the segment A of the superior graphic. At 107, the portions of the array represented by the contents of TEMP3 and TEMP4 are deleted. It should be noted that many of the foreground superior pels will be superposed over pels of the superior graphic as well on occlusion pels of adjacent superior graphic pels 50, all as shown in Figure 5. At 108., the resultant occlusion mask is NOT ANDed with the inferior graphic to delete those portions of the inferior graphic which coincide with the original foreground pels of the superior graphics and the occlusion pets added for generating the contoured separation space. Then at 109, the new graphic NG is generated by logically OR combining the A segment of the superior graphic with the inferior graphic fragmented pattern generated at 108. Then the new graphic is stored at 110 with the remainder of the superior graphic to provide a composite graphic resulting from two original graphics. Original graphics can be preserved in their own respective data storage spaces while the new graphic is assigned to and stored in a new area of DASD 25.

Figure 9 shows processing elements used to effect the machine-implemented process of Figure 8. A data storage unit 120 stores a superior graphic raster pattern at 121 and an inferior graphic raster pattern at 122. Both of the raster patterns have the same relative address space identified as an array of pels (binary ones and zeroes) having a reference point 0 at the upper left-hand corner of the respective patterns. The horizontal ordinate extends to location X while the vertical ordinate extends downwardly to location Y. The inferior graphic pattern (foreground pels) is measured and the desired separation J (on integral numbers of pcl positions) inputted by the user at 135 is added to the measured inferior graphic pattern stored at 122 resulting in upper and lower boundaries M and N along the vertical ordinate 0-Y of array 122. M and N correspond to double-headed arrow 88 of Figure 6. The graphic pattern extent to be processed by each of the arrays 121 and 122 is defined at 125 as being MN for the vertical ordinate with the horizontal ordinate extending from 0 to X. The portion M to N plus "j" borders (shown as M+j and N+j) of superior pattern array 121 is read at 126 into array registers A 127. Registers 127 now contain that segment of the superior graphic pattern used in the superposing machine processing. At 128 and 129 TEMP4 and TEMP3 are added for accommodating the array processing steps of vertical and horizontal shifts. Element 136 is the control which responds to the user inputted space J for defining the main memory data storage areas 128 and 129. The resultant occlusion array (viz array 71 of Figure 5) moves over line 142 to NOT AND segment 140 which also receives the inferior graphic pattern .between extents M and N from the array 122 for NOT ANDing the two patterns together. NOT AND 140 deletes those foreground pels of the inferior graphic patterns which coincide with the occlusion pels of the expanded A array (as shown in Figure 4). At TEMP2 (a portion of the main memory ₁₂₀) the fragmented array corresponding to element 73 of Figure 5 is stored. That array is combined logically at 147 with the superior graphic array 121 to create the new graphic pattern NG which is presented on the display 11 as. pattern 148. Alternately, the new graphic pattern can be stored within data storage unit or main memory 120 for relaying to DASD 25.

Figure 10 shows the APL flowchart for implementing the control 136 for creating the occlusion array from the superior graphic pattern for the desired contoured separation space TTC. The control is entered at 150 from other controls which are represented in the drawings and description as being data transfer and array selection (addressing) procedures which are known machine operations and hence not detailed. The superior graphic expansion steps started at 150 include setting up control signals for subsequent iterative loops at 152, 155, 157. The parameter Z controls the vertical aspects of creating occlusion pels. Z=1 indicates occlusion pels are being added above the horizontal ordinate while Z=0 indicates occlusion pets are being added below the horizontal ordinate. The parameter "X" controls the lateral or horizontal aspects of creating occlusion pels. The parameter X=1 indicates that the left-hand semicircle of circle 51 is to be processed; X=0 indicates the right-hand semicircle is to be processed. In an internal computer regisier (not shown) designated as L7, the parameter I is set to zero. I is the count control for indicator for left- and right-hand lateral movements as will become apparent; shown in the Figure 3 in the horizontal expansion of a single pel 50. Internal register L9 (not shown ) has the parameter K equal to zero. K is the count control for vertical aspects as indicated by the arrows ₅2, 53, and 54 of Figure 3. Numerals 152, 155, 157 designate the return points for all loops in the pel expansion or occlusion pattern generation process. It should be realized that the Figure 10 illustration results in processing the step 161 is again re-executed but now Z is equal to zero indicating that all I of the occlusion pels represented by arrow 52 have been added to the segment A. At 153, the pel positions lying on the pel 50 ordinate are determined by the values of I and X. "I" is incremented by one at step 163. Then at 1611, the value I is compared with TTC. Since "I" had been preset to zero for identifying the vertical ordinate intersecting the supergraphic pel 50, it is certainly not greater than TTC. Accordingly, at 159, the current ordinate position is moved to the left by one pel position and a new value TEMP is calculated for the new ordinate position as in step 154; steps 155-162 are repeated for generating the occlusion pels represented by the left-most arrow 53 of Figure 3. This procedure is repeated until the horizontal parameter I exceeds the desired contoured spacing TTC at step 164, at which time at step 166, APL module 35. checks the hemispheric parameter X for unity (as set at 150). If X is unity, then only the left hemisphere of the circle 51 has received occlusion pels. Then at 167, at half-loop calibration step sets 1=0, Z=1, and X=0. The above-described steps beginning at 152 are repeated for the right-hand hemisphere of circle 31. Next time, machine step 166 is executed, X=0 indicating that all of the occlusion pels have been added for the superior graphic pels within the respective circle 51 indicating that the expanded superior graphics array for masking or occluding the inferior graphic pattern has been completed within. Remember the Figure 10 illustrated steps occur for all segment A foreground pels as known for APL array processing. Accordingly, at 168, the APL module 35 proceeds to step 107 for deleting the extra areas, and step 107A for NOT ANDing the mask with the inferior graphic pattern.

While all of the graphic patterns illustrated in the drawings have been at relatively coarse resolutions for the entire superior graphic array of segment A. Refer momentarily to Figure 3, when arrow 52 shows adding a single occlusion pel for superior graphic pel 50, in the constructed implementation all I pels within the superior graphic will have an occlusion pel added in the same relative position in the expanded superior graphic or occlusion array. At 151, an array (TEMP2) of zeros or background pel indicating signals is created that has the dimensions of segment A. TEMP2 accumulates at 158 the occlusion pels created by the above-described expansion of segment A foreground pets. At 158, an occlusion pel is added to each pel of the segment A of the superior graphic corresponding to the first pel position above superior graphic pet 50 of Figure 3. The value of K is incremented by one at step 159. Then, at step 160, the value of K is compared with a value TEMP created at step 154, i.e., have the number of occlusion pels been added vertically above the superior graphic pel 50 equal to the desired contoured space. If not, then the described process repeats at the next vertical pel position, i.e, two pets above superior graphic pel 50, for example. At 158, additional occlusion pets are added and the loop repeated until at 160 the value K exceeds the value TEMP. At this point the upwardly moving arrow 52 of Figure 3 has been completed for all of the graphic pels (binary ones) of the segment A. Please note that background pels result in no occlusion pels being added.

From the loop 157-160, the APL superpose module 35 proceeds to step 161 which determines whether or not processing has been completed for the downward arrow 52, i.e., is the parameter Z equal to zero; if not, only the occlusion pels above the foreground pel 50 have been added. Accordingly, at 162 the parameter Z is set to zero and junction point 155 returned to for repeating steps 157-160 for generating the occlusion pets below supergraphic foreground pel 50. Then patterns used to illustrate contouring a separation space between a set of graphic patterns being superposed, it should be understood that at high resolution graphics (240 pels per inch and greater) that the practical effect of first expanding the inferior graphic pattern for segmenting the superior graphic pattern; then expanding the segmented (or complete superior graphic pattern) for generating a mask or occlusion array followed by NOT ANDing the occlusion array with the inferior graphic pattern and then reconbining the fragmented array, (resulting from the NOT ANDing the inferior graphic with the original superior graphic) results in a set of superposed graphics separated by a contoured space to the superior graphic having a typographic quality contoured spacing selected by the user. The advantage of the selective spacing is that the spacing can be selected in accordance with the type of graphics being superposed. For example, with character graphics there are fonts that have relatively heavy weight characters and fonts that have relatively light weight characters. The terms "heavy weight" and "light weight" refer to the width of the strokes or lines constituting the characters. For a heavy character, a greater spacing is desired while for a light weight character, the spacing can be substantially less for obtaining typographic aesthetics. It should also be borne in mind that many fonts have -character graphics with serifs and other ornate graphic configurations which the above-described. processing accommodates for ensuring a contouring of the fragmented or intercepted inferior graphic to the superior graphic including having pels of the inferior graphic in the interior background portions of the superior graphic. Such contouring not only automatically follows the general contour of the superior graphics, but also matches the fineness of the design of the superior graphics, all without human intervention or observation.