GB1598342A - Display systems - Google Patents

Description

(54) IMPROVEMENTS IN OR RELATING TO DISPLAY SYSTEMS (71) We, INTERNATIONAL COMPUTERS LIMITED, a British Company, of ICL House, Putney, London, S.W.15, do hereby declare the invention for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement: This invention relates to display systems, and in particular to systems for displaying grahics data on a display device which forms the display by repeatedly scanning a plurality of picture elements in a predetermined sequence. The most common current example of such a device is the raster-scanned cathode-ray tube. Graphics data includes figures or diagrams of a kind which cannot be displayed by a purely alphanumeric display system, together, possibly, with textual matter.

A graphic display will normally contain a number of graphical structures (for example geometrical figures) which are intended to be seen by the user, and the data to be displayed is often generated or supplied in a form including definitions specifying these structures in a manner related to their overall shape or characteristics. This form of representing the data is not, as it stands, suitable for driving a display system such as a raster-scanned cathode-ray tube, which must be supplied a drive signal correlated with the order in which the picture elements are scanned rather than the arrangement of the structures to be displayed.

The process of deriving the drive signal for the display device from the original description is sometimes referred to as "scan conversion" and a number of proposals have hitherto been made as to how it is to be carried out.

This invention provides a system for driving a display device to cause it to produce a display formed of plurality of picture elements which are repeatedly scanned in a predetermined sequence, the system comprising: means arranged to hold in operation a first compressed representation of the display, which representation comprises at least one separately identifiable definition of a graphical structure to be seen in the display; means arranged to expand the first compressed representation into an expanded representation in which there is a separate information item for each picture element of the display, which item comprises an indication of the display value of that picture element;; means arranged to compress the expanded representation into a second compressed representation of the display, in which representation the display values of the picture elements of the display are defined in the order in which they are scanned; means arranged to store the entire second compressed representation; and means arranged repeatedly to output the second compressed representation and, each time it is output, to decompress it to produce a drive output for the display device for controlling the display value of each picture element in turn as it is scanned in the said predetermined sequence.

It also provides a method of driving a display device to cause it to produce a display formed of a plurality of picture elements which are repeatedly scanned in a predetermined sequence, the method comprising expanding a first compressed representation of the display, which representation comprises at least one separately identifiable definition of a graphical structure to be seen in the display, into an expanded representation of the display in which there is a separate information item for each picture element of the display, which item comprises an indication of the display value of that picture element, compressing the expanded representation into a second compressed representation of the display, in which the display values of the picture elements are defined in the order in which they are scanned, storing the second compressed representation, and repeatedly decompressing it to produce a drive output supplied to the display device, which drive output controls the display value of each picture element in turn as it is scanned in the said predetermined sequence.

Preferably, expansion takes place in stages in each of which information items relating to some only of the picture elements of the display are produced, compression of data produced elements of the display are produced, compression of data produced in any of the stages into the second compressed representation starting before the start of the next expansion stage (if any).

An example of a display system embodying the invention will now be described in greater detail by way of example with reference to the accompanying drawings, in which: Figure I is a block diagram of the system; Figure 2 is a diagram showing how the display data is represented at various points in the system; and Figure 3 is a more detailed block diagram of the decompressing unit of the system.

Referring to Figures 1 and 2, the system is intended to display primarily graphical data on a standard TV monitor 1 which is repeatedly scanned in a raster pattern. An example of a display produced by the monitor is shown in Figure 2a as a display 2. The individual scan lines, such as line Yr, are formed of separate picture elements 3, and the monitor allows the display value of each element to be individually controlled. It will be assumed for the present that the possible display values are different colours. In Figure 2 the display 2 is of a quadrilateral of solid colour having vertices with coordinates xl Yi to x4 y4.

Information to be displayed is transmitted over a path 4 from an external processor, which may be a computer mainframe, either local or at the remote end of a communications link, or a satellite processor. The path 4 is connected to an interface unit 5 of a microprocessor 6. The microprocessor 6 has a highway 7 to which the interface unit 5 is attached, as are the other units of the microprocessor: a central processing unit 8; a programmable read-only memory unit (PROM) 9; a random-access memory unit (RAM) 10; and a second interface unit 11.

The PROM 9 includes a section 91 which controls the operation of the microprocessor 6.

A section 101 of the RAM 10 is used for working storage by this and other sections of the PROM 9.

Graphical data received over the path 4 is stored in the form in which it is received in a section 102 of the RAM 10. This section holds a complete display file containing definitions of graphical structures to be displayed. The structures may be geometrical objects which are of solid colour or are constructed of unit thickness lines. Objects of solid colour are, in this example, represented by the vectors which define the outline of the object together with an indication of the colour of the interior of the object. Thus Figure 2b shows the information defining the quadrilateral of the display 2 in Figure 2a. It is supplied as starting co-ordinate xl, yl, and the increments Vexs, Ay1, to Ax4, Ay4 to successive vertices forming the boundary vectors.Associated with the object and following its initial co-ordinates is a display-value code d1. In the system being described the display value code has three bits, each defining whether one of the red, blue or green guns should be on or off. That gives eight colours in all (including black).

Further graphical information can be input via the interface unit 11 from a device 12 such as a light pen used interactively with the display. Fixed manipulative routines in a section 92 of the PROM 9 merge the information into the display file 102. The information is also supplied over a line 13 to the external processor.

The coded graphical information in the display file 102, which is oriented to the structures to be displayed, and is highly compressed since only relatively few points need be specified to define the structure, is then converted into a form suitable for driving the monitor 1. It is first expanded. Fixed routines in a section 93 of the PROM 9 map one zone of the display at a time into a section 103 of the RAM 16. For this purpose the display is divided into equal non-overlapping areas, one of which is shown bounded by the scan lines Ym and Yn in Figure 2a. The zone that is then mapped (shown bounded by dotted lines in Figure 2a) is this area plus the preceding scan line Ym~l, which is mapped twice, once in this zone and once in the preceding one. In the section 103 a separate item of information 1031 (see Figure 2c) for each picture element in the zone is held in a predetermined location. The item, at the end of the mapping process, holds the display value of the element, as shown. It also contains a further two bits (not shown) which are used in the mapping process, as will be explained.

For each zone the mapping routines 92 pass through the whole display file 102 and use a clipping technique to determine if the vectors of each object in turn intersect the zone. Each of those that do is mapped by setting to 1 one or other of the extra bits of each element on the vector. The first is set for each element which lies on the boundary of a solid object to either its left or right. If the element does not have such a neighbour, as for example, the vertex X4, y4, the second extra bit is set. Horizontal boundaries need not be mapped, provided their end-points are included in the adjacent vectors, although they can also be mapped by setting the second bits of the elements of the boundary. The second bit is also set for unit-thickness lines, whether open-ended or closed.

When the vectors defining an object have been written in, the mapping routines trace along each row of display elements in the zone in turn, checking the two extra bits of each element. When they first meet a set first bit they set a flag to indicate that a solid object has been met and write the display value of the object into the information item of that element. They then fill in the interior by writing the display value into each subsequent element while the flag is still set. They continue to check the extra bits and when a set first bit is again met (indicating a closing boundary) they write the display value of the object into that element and then clear the flag. The display value of the object is not written into subsequent elements unless a set first bit is again met, when a further segment of that display value will be written.This allows re-entrant objects or objects with hollows to be created.

At the same time each second bit is checked and, if it is set, the object's display value is written into just the item of the element concerned. This completes the vertices or horizontal boundaries of a solid object, or alternatively draws in a line object.

As soon as either of the two extra bits has been found set and acted on it is cleared to 0.

Thus when the object has been fully mapped there will be no set extra bits. The display value of the object overwrites the previous values, so that the object overlies any previously mapped objects. But any previously mapped objects or parts of objects outside the area of the one currently being mapped are not affected, since they have no set extra bits.

The first item in the display file is a background colour; if there is no such entry the colour is assumed to be black. In Figure 2c the background is shown as do.

Mapping is continued until all objects in the zone have been mapped. When the mapping of a zone is complete fixed compression routines in a section 94 of the RAM 9 compress the data of that zone into a coded version shown in Figure 2d and transmit it to a display driver 14. The routines basically specify the points, taking them in the order of scan of the monitor 1, at which there is a transition to a new display value, detected by comparing each stored display value with that of the previous element.

Transitions are represented in two ways. When a transition is not related to a transition on the previous scan line, as at the point x2 Y2, two code elements are used. The first specifies the display value of the first picture element after the old display value (the transitional element). The second specifies the length of run, starting with the next element (that is, the one after the transition element) and continuing up to the next transition. Thus the line Yr is coded as rl dl r2 do r3 where rl = x2 - 1, r2 = x3 - x2 and r3 = N - x3 - 1, N being the total number of elements in a line. When, as here, no initial display value is specified it is assumed to be do. The final code may instead be a new-line code.

But many transitions are related to a transition in the previous line because they lie on the same line or curve. Provided they fall within a predetermined distance of the transition in the previous line they are represented by a correlation code specifying the relationship.

Thus the line y, is represented by a code CL showing that the first transition is one ot the left of the transition above; a code C,. showing that the next transition is vertically below the following transition in the line above; and a new-line code n. The same coding sequence applies to each remaining line of the mapped zone. No display value code is needed, since the colours are the same as in the line above.

The line Ym~l is mapped in this zone to allow correlations in the first new line of the zone, Ym, (if they exist) to be determined.

The data represented by the mapped zone is transferred to a compressed-data store 15 in the display driver 14 (see Figure 3). The next zone is then mapped into the same section 103 of the RAM 10 as the last, compressed and transferred to the store 15 where it is added to the data already transferred. When the whole display file 102 has been dealt with the store 15 holds the compressed data for the complete display.

To refresh the monitor this data is output repeatedly once per scan, and decompressed dynamically. The store 15 has an address register which is incremented by a signal from a decoder and control unit 16. The stored code elements are thus retrieved in succession. The unit 16 decodes the code element and if it detects a run-length code passes it to a look-up dictionary, a read-only memory 17 which in response to the run-length code as an address outputs the binary value of the run. This value is loaded into a counter 18. The colour code for the run is already present at this stage in a three-bit colour register 19, having been introduced, as will be explained, in response to one of the other two code types.

A sync unit 20 outputs the usual sync pulses to control the monitor and also clock pulses synchronised with the picture elements as they are scanned. The counter 18 is decremented by these pulses and when it reaches zero issues a signal to the decoder and control unit 16 that the run is completed. The unit 16 then calls for the next code element.

During the run the value held in the colour register is entered into a line store 21, a shift-register three bits wide (one for each colour), having one more stage than there are picture elements in a scan line, and clocked at the picture element rate. The colour value is entered via a gating unit 22.

The contents of the line store 21 are output from the second to last stage as red, green and blue signals to control the monitor. Runs are thus entered into the line store one complete scan line ahead of being supplied to the monitor.

When a correlation code is received it is decoded to control the gating unit 22, which is constructed of standard combinational logic elements. It causes the output of the line store 21 to be recirculated so that it can be output again during the following scan line. Outputs from the three last stages of the line store 21 are supplied on paths 23, 24 and 25 to the gating unit 22, which selects the appropriate output in accordance with the particular correlation code. If the correlation is such that the transition is vertically below that in the previous line the output on the path 24 is selected and re-enters the line store in the same position along the scan line as it was output. If it is a left correlation the output on the path 23 is selected and is consequently advanced by one element.If it is a right correlation the output on the path 25 is selected and is delayed by one element. The output of the colour register 19 is inhibited during the recirculation by the gating unit 22.

The values output from the gating unit 22 are also transmitted to the decoder and control unit 16. There each is compared with the previous value. Inequality indicates that a transition to a new colour has taken place. The unequal value (corresponding to the element after the transition) is entered into the line store 21 and used to set the colour register 19 for the next run or correlation. Then the next code element is decoded.

Transition codes are entered into the colour register 19 and then the line store 21.

The colour values output by the line store 21 are passed to a video unit 26 which converts the logic signals output by the line store 21 to a form suitable for applying to the guns of the monitor 1. It is connected to the highway 7 and can receive mode signals causing it to direct the output of the line store to a permanent-copy device such as a plotter or microfilm printer.

In some circumstances it may be convenient to map the geometrical information and compress it in the external processor. The compressed data is then transmitted directly to the compressed data store 15 from the interface 5.

The interface 5 may either receive data in serial, communications, format, or in parallel format for direct memory access, depending on the location of the external processor.

The programming of the PROM 9 is by well-known programming techniques and is arranged to carry out the operations described.

As an example, the system described may produce a display of 512 by 512 picture elements each having eight possible colours defined by a three-bit code. The display may be divided into 32 zones of 16 lines (excluding the line replotted from the preceding zone) and the compressed-data store 15 may have a capacity of 96K. If the entire display were mapped on a point-by-point basis the zone map would require 768K without, and 1300K with, the extra bits used for the mapping; in the system described the mapping takes place in 43K, which may be used for other purposes when mapping is not taking place. The compressed-data store will be sufficient for a display which can be compressed at a compression ratio of 8:1 or better; in practice this is ample for all ordinary graphical displays.

The display value may, of course, instead of a colour, represent a grey-scale (brightness) value on a black-and-white monitor, or be a colour but have more bits to allow different brightness in the primary colours.

Various modifications may be made.

Instead of mapping and compressing the display values themselves, the display value information stored for each picture element may be a reference to a value stored in a table held in the video unit 26 which is written from the highway 7 and defines the signals passed to the gun or guns. This allows a selection to be made from a wider range of display values, although at any one time only as many may be used as the total number of reference values allows.

The display value may be a Z value for three-dimensional applications. In that case, fresh Z values may be input over the line 4 and compared with the existing values in the zone map 103, overwriting them if they are greater (nearer the viewer). This allows hidden surfaces to be eliminated. The zone map 103 may either be recreated from the display file 102 for this purpose or reconstructed by recycling the output of the line store 21 via the unit 26 and highway 7 to the zone map 103.

Interlace may be provided by compressing in two successive fields, correlation being with the proceding line of the same field. Mapping of the complete display may then take place twice, once for each field and by alternate lines.

Instead of using extra bits to indicate boundaries of objects, vectors may be plotted directly by writing the display value of the object into the items of the elements concerned.

This reduces the store required for the zone map but increases the complexity of the routines required to fill in the interior of a solid object.

In an alternative way of representing the original graphical information the display file holds structures consisting of a vector together with a display value. That display value is then written into the elements of the vector and all elements to their right. Solid or unit-thickness objects are then created by specifying subsequent vectors to the right of the original one which overwrite some of the just-written elements.

The display driver described uses a simple coding scheme in which correlation codes provide a single correlation, either vertically, or by one element to left or right. More complicated correlation relationships may be dealt with by employing the coding scheme and decoding apparatus described in our copending British Patent Applications 14122/77 Serial No. 1598343 and 14123/77 Serial No. 1598344.

The coding scheme described uses separate transition and run-length codes. This has the advantage of avoiding the need to transmit display-value and run-length information simultaneously. But if desired the run-length codes may encode the full length of the run (rather that the length excluding the initial element) and contain the display value.

WHAT WE CLAIM IS: 1. A system for driving a display device to cause it to produce a display formed of plurality of picture elements which are repeatedly scanned in a predetermined sequence, the system comprising: means arranged to hold on operation a first compressed representation of the display, which representation comprises at least one separately identifiable definition of a graphical structure to be seen in the display; means arranged to expand the first compressed representation into an expanded representation in which there is a separate information item for each picture element of the display, which item comprises an indication of the display value of that picture element;; means arranged to compress the expanded representation into a second compressed representation of the display, in which representation the display values of the picture elements of the display are defined in the order in which they are scanned; means arranged to store the entire second compressed representation; and means arranged repeatedly to output the second compressed representation and, each time it is output, to decompress it to produce a drive output for the display device for controlling the display value of each picture element in turn as it is scanned in the said predetermined sequence.

2. A system as claimed in Claim 1, in which the arrangement is such that expansion takes place in stages, in each of which information items relating to some only of the picture elements of the display are produced, compression of data produced in any of the stages into the second compressed representation starting before the start of the next expansion stage (if any).

3. A system as claimed in either of the preceding claims, in which the second compressed representation comprises run-length encoding.

4. A system as claimed in any one of the preceding claims, in which the second compressed representation comprises correlation encoding between sequences of picture elements.

5. A system as claimed in both claim 2 and claim 4, in which the arrangement is such that some picture elements are each represented by an information item in each of two stages, one used in generating the part of the second compressed representation defining the display value of the picture element represented by that item and the other serving to allow correlation encoding in the second compressed representation of picture elements other than that represented by the item.

6. A system as claimed in any one of the preceding claims, in which the means arranged to expand the first compressed representation comprises a control store, a processor and means arranged to hold (in its entirety or in stages) the expanded representation, all connected to a common highway, and the means arranged to hold the first compressed representation is also connected to the common highway.

7. A display system comprising a system as claimed in any one of the preceding claims, and a display device arranged to receive the said drive output.

8. A method of driving a display device to cause it to produce a display formed of a plurality of picture elements which are repeatedly scanned in a predetermined sequence, the method comprising expanding a first compressed representation of the display, which

**WARNING** end of DESC field may overlap start of CLMS **.

Claims (14)

**WARNING** start of CLMS field may overlap end of DESC **. Interlace may be provided by compressing in two successive fields, correlation being with the proceding line of the same field. Mapping of the complete display may then take place twice, once for each field and by alternate lines. Instead of using extra bits to indicate boundaries of objects, vectors may be plotted directly by writing the display value of the object into the items of the elements concerned. This reduces the store required for the zone map but increases the complexity of the routines required to fill in the interior of a solid object. In an alternative way of representing the original graphical information the display file holds structures consisting of a vector together with a display value. That display value is then written into the elements of the vector and all elements to their right. Solid or unit-thickness objects are then created by specifying subsequent vectors to the right of the original one which overwrite some of the just-written elements. The display driver described uses a simple coding scheme in which correlation codes provide a single correlation, either vertically, or by one element to left or right. More complicated correlation relationships may be dealt with by employing the coding scheme and decoding apparatus described in our copending British Patent Applications 14122/77 Serial No. 1598343 and 14123/77 Serial No. 1598344. The coding scheme described uses separate transition and run-length codes. This has the advantage of avoiding the need to transmit display-value and run-length information simultaneously. But if desired the run-length codes may encode the full length of the run (rather that the length excluding the initial element) and contain the display value. WHAT WE CLAIM IS:

1. A system for driving a display device to cause it to produce a display formed of plurality of picture elements which are repeatedly scanned in a predetermined sequence, the system comprising: means arranged to hold on operation a first compressed representation of the display, which representation comprises at least one separately identifiable definition of a graphical structure to be seen in the display; means arranged to expand the first compressed representation into an expanded representation in which there is a separate information item for each picture element of the display, which item comprises an indication of the display value of that picture element;; means arranged to compress the expanded representation into a second compressed representation of the display, in which representation the display values of the picture elements of the display are defined in the order in which they are scanned; means arranged to store the entire second compressed representation; and means arranged repeatedly to output the second compressed representation and, each time it is output, to decompress it to produce a drive output for the display device for controlling the display value of each picture element in turn as it is scanned in the said predetermined sequence.

2. A system as claimed in Claim 1, in which the arrangement is such that expansion takes place in stages, in each of which information items relating to some only of the picture elements of the display are produced, compression of data produced in any of the stages into the second compressed representation starting before the start of the next expansion stage (if any).

3. A system as claimed in either of the preceding claims, in which the second compressed representation comprises run-length encoding.

4. A system as claimed in any one of the preceding claims, in which the second compressed representation comprises correlation encoding between sequences of picture elements.

5. A system as claimed in both claim 2 and claim 4, in which the arrangement is such that some picture elements are each represented by an information item in each of two stages, one used in generating the part of the second compressed representation defining the display value of the picture element represented by that item and the other serving to allow correlation encoding in the second compressed representation of picture elements other than that represented by the item.

6. A system as claimed in any one of the preceding claims, in which the means arranged to expand the first compressed representation comprises a control store, a processor and means arranged to hold (in its entirety or in stages) the expanded representation, all connected to a common highway, and the means arranged to hold the first compressed representation is also connected to the common highway.

7. A display system comprising a system as claimed in any one of the preceding claims, and a display device arranged to receive the said drive output.

8. A method of driving a display device to cause it to produce a display formed of a plurality of picture elements which are repeatedly scanned in a predetermined sequence, the method comprising expanding a first compressed representation of the display, which

representation comprises at least one separately identifiable definition of a grahical structure to be seen in the display, into an expanded representation of the display in which there is a separate information item for each picture element of the display, which item comprises an indication of the display value of that picture element, compressing the expanded representation into a second compressed representation of the display, in which representation the display values of the picture elements are defined in the order in which they are scanned, storing the second compressed representation, and repeatedly decompressing it to produce a drive output supplied to the display device, which drive output controls the display value of each picture element in turn as it is scanned in the said predetermined sequence.

9. A method as claimed in claim 8, in which expansion takes place in stages, in each of which information items relating to some only of the picture elements of the display are produced, compression af data produced in any of the stages stage into the second compressed representation starting before the start of the next expansion stage (if any).

10. A method as claimed in either claim 8 or claim 9, in which the second compressed representation comprises run-length encoding.

11. A method as claimed in any one of claims 8 to 10, in which the second compressed representation comprises correlation encoding between sequences of picture elements.

12. A method as claimed in both claim 9 and claim 11, in which some picture elements are each represented by an information item in each of two stages, one used in generating the part of the second compressed representation defining the display value of the picture element represented by that item and the other seving to allow correlation encoding in the second compressed representation of picture elements other than that represented by the item.

13. A method as claimed in any one of the claims 8 to 12 in which each of the said information items includes non-display-value information, and during the expansion characteristics of the display are introduced into the items as non-display-value information, and thereafter display-value information is introduced into the items of elements determined in response to non-display-value information.

14. A system substantially as hereinbefore described with reference to and as shown in the accompanying drawings.