Background of the Invention

Technical Field

This invention relates in general to computer graphics display systems, and more particularly, to a processing method for identifying a displayed object that intersects an operator selected area of the display screen and for enlarging a selected area of the display screen to facilitate operator selection of an object when a complex display it presented.

Description of the Prior Art

Interactive raster graphics systems, such as CAD/CAM workstations, are widely used to design components and systems of mechanical, electrical, electromechanical and electronics devices. Frequently, the emphasis within such systems is on an operator's interacting with a computer based model of a component or system being designed in order to test, for example, its mechanical, electrical or thermal properties. The computer based model is comprised of numerous graphics objects that are individually processed and displayed for operator action. Selection of a display object is accomplish­ed via any one of a number of operator controlled interaction devices, such as light pens, locators (e.g., a data tablet with stylus), and alphanumeric and function keyboards. An important part of many interaction sequences is computer identification of an operator selected displayed object to be operated upon, a process which is complicated by the pipeline processing techniques utilized within most graphics systems today.

Typically, graphics system processing techniques require the reprocessing of an entire display screen to identify a particular displayed object selected or picked for further processing. More particularly, existing processing methods require the reexecution of the whole display program, includ­ing: transformation of each geometric primitive defining a displayed object in world coordinate space; clipping of each transformed primitive against the predefined clipping boundary in world coordinate space; mapping of each clipped primitive to an operator defined viewport in screen coordinate space; rasterization of all mapped data; and finally, determination of whether the generated pixels intersect the operator defined selection area or window in screen coordinate space. If so, then a pick occurs.

Although effective, a clear drawback to the procedure is the unnecessary processing occuring as the result of mapping and rasterizing geometric objects not selected. For example, rasterization of a filled polygon outside the operator select­ed screen area is both unnecessary and time consuming. A similar processing technique is also disadvantageously used in most graphics systems to enlarge or "zoom out" an operator selected area of a complex display screen to facilitate selection of a particular displayed object. Again, rerasteri­zation of objects in nonselected areas of the display screen ultimately comprises unnecessary processing. Thus, a genuine need clearly exists for a more efficient interactive raster graphics display system processing technique for identifying an operator selected displayed object to be operated on and for zooming out an operator choosen area of the display screen to facilitate picking of a displayed object.

Summary of the Invention

The present invention is defined in the attached claims.

Briefly, in a computer graphics display system having means for mapping data representative of geometric primitives in a first coordinate space to a second, planar coordinate space for display, this invention comprises an improved method for identifying displayed primitives that intersect an opera­tor selected area of the second coordinate space. The method comprises the steps of: inverse mapping the second coordinate space selected area to first coordinate space; clipping the data list representative of displayed geometric primitives against the inverse mapped selected area; and selecting the clipped data representative of displayed geometric primitives that intersect the inverse mapped selected area in the first coordinate space.

In a further enhancement, the invention comprises a processing method for enlarging displayed primitives that intersect an operator selected area of the second coordinate space to facilitate operator selection of a particular dis­played object. The method includes the steps of: inverse mapping the second coordinate space selected area to the first coordinate space; clipping the data list representative of displayed geometric primitives against the inverse mapped selected area in the first coordinate space and discarding data outside the inverse mapped area; mapping the clipped data that intersects the inverse mapped selected area to a zoom viewport within the second coordinate space; and rasterizing the mapped data for display within the second coordinate space zoom viewport. The identification processing method outlined above can then be combined with this aspect of the invention to identify an operator selected displayed object within the zoom viewport.

Accordingly, a principal object of the present invention is to provide a graphics system identification processing method capable of efficiently and quickly evaluating displayed data to identify those geometric primitives that intersect an operator selected area of the display screen.

Another object of the present invention is to provide such an identification processing method wherein only primi­tives that intersect the operator selected area of the display screen are mapped to and rasterized in screen coordinate space.

Yet another object of the present invention is to provide an improved graphics system processing method capable of efficiently and quickly zooming out a selected area of the display screen to facilitate selection of a particular geo­metric object from a complex screen display.

Still another object of the present invention is to provide such a zoom processing method wherein only objects that intersect the operator selected area of the display screen are mapped to and rasterized in screen coordinate space.

Brief Description of Drawings

These and other objects, advantages and features of the present invention will be more readily understood from the following detailed description of one embodiment of the present invention, when considered in conjunction with the accompanying drawings in which:

• Figure 1 is a general block diagram representation of an interactive computer graphics display system incorporating the present invention;
• Figure 2 is a flow diagram representation of existing graphics system processing used to identify an operator selected displayed object;
• Figures 3A and 3B graphically illustrate certain areas defined in world coordinate space and screen coordinate space with implementation of the present invention;
• Figure 4 is a flow diagram representation of the selec­tion processing method of the present invention;
• Figure 5 is a flow diagram representation of the corre­sponding zoom processing technique of the present invention; and
• Figures 6A and 6B graphically illustrate certain areas defined in world coordinate space and screen coordinate space with implementation of the method of the Figure 5.

Detailed Description of the Invention

This invention comprises an improved interactive raster graphics display system processing method for identifying an operator selected displayed object which requires rerasteri­zation of only a limited portion of the original display data, thus dramatically reducing processing time. Further, the invention comprises an analogous graphics system processing technique for zooming out or enlarging objects displayed in a limited portion of the monitor screen surrounding the center of a locating device, to facilitate operator selection of a particular displayed object when a complex display is pre­sented. Only the contents of a "zoom window" defined in world coordinate space are rerasterized, again significantly reduc­ing processing time over prior zoom techniques.

Referring to Figure 1, several major components of a raster graphics display system, generally denoted 10, are illustrated. System 10 includes system memory 12, display processor 14, pixel processor 16, video pixel memory 18, monitor 20 and selector device 22. Briefly described, display processor 14 is responsible for executing the display program residing within system memory 12 and functions principally to generate the image which will appear on display monitor 20. Display processor 14 performs the following functions: decoding graphics orders and executing nondrawing orders, e.g., bookkeeping and control; performing transformation and clipping functions to geometric primitives, such as lines, characters, polygons, etc.; preparing geometric objects or primitives for display by preprocessing and feeding data representative thereof to pixel processor 16 and video pixel memory 18; and if an operator selection or pick occurs, processing the selected data and sending it back to the application program. Pixel processor 16 includes a hardware implementation of the Bresenham Line Generation Algorithm, which takes the end points of a vector as input (i.e., x-axis and y-axis coordinates) and generates pixels in video pixel memory 18 as output for display on monitor 20. Textual treatment of the Bresenham Algorithm is provided by J. D. Foley and A. Van Dam, in Fundamentals of Interactive Graphics at pp. 422-461 (1982). Processor 16 also contains hardware for pixel manipulation, such as logical operations, bit block transfer, etc. Video pixel memory 18 consists of eight 1k by 1k bit planes which support simultaneous display of up to 256 colors via color look-up tables. The image stored in video pixel memory 18 is displayed in monitor 20 for operator viewing. Selector device 22 is connected to processor 14 and allows direct operator interaction with system 10.

As noted initially, several types of selector devices are widely in use, such as locators, light pens, valuators, keyboards and buttons. Of these, locators comprise the vast majority of such devices, tablets being perhaps most common. A tablet is a flat surface over which a stylus (like a pencil) or hand cursor is moved. The position of the hand cursor or stylus is transferred to processor 14 of system 10. Most cursors also provide the operator with several function keys, such as the "pick" and "zoom" functions to which the present invention is directed. Actuation of a function key directs the processor go begin the desired processing with reference to the current location of the selector device. Tablets are well known in the art and numerous versions are commercially available.

As briefly discussed above, previous picking procedures have required the rerasterization of an entire screen display to identify a particular geometric object selected for appli­cation processing. This prior art procedure is illustrated in the flow diagram of Figure 2. Actuation of the "pick" func­tion key enables the pick correlation operation at 40 "Start", and processor 14 is directed to obtain for processing the next piece of data representative of a geometric primitive, "If Not End Of Data, Fetch Data" 42. Processor 14 then transforms and clips the data in world coordinate space "Transform Geometric Data" 44, and "Clip Data Against Operator-Defined Clip Boundary" 46, respectively. (Transformation, clipping and mapping functions are described further below.) Clipping instruction 46 essentially directs the processor to remove those parts of a displayed object, or more particularly, one or more geometric primitives definitive thereof, not within a defined clipping volume (3D clipping) or area (2D clipping). After clipping, processor 14 is directed to map the clipped data to a predefined display viewport in screen coordinate space, "Window-to-Viewport Mapping" 48, and then to rasterize the mapped object, "Rasterize Mapped Object" 50. After rasterization, the display processor is directed to finally determine whether the pixels generated during rasterization intersect the operator defined pick window or area in screen coordinate space, "Check For Pick And Log Picked Data" 52. If an intersection is identified, a pick occurs and the display processor records the object for subsequent processing accord­ing to the application rules set by the operator.

Again, rasterization of mapped objects not subsequently selected results in a time consuming and unnecessary expen­diture of processing time. Thus, the present invention comprises a processing technique whereby only those geometric objects intersecting the pick area in screen coordinate space are ultimately rasterized from world coordinate space to screen coordinate space. Before describing a detailed imple­mentation of the processing technique, standard transforma­tion, clipping and mapping functions carried out by graphics system 10 will be briefly reviewed.

With the present invention, each primitive geometrically oriented in a 3D or 2D world coordinate space (such as a vector, circle, curve or filled polygon) representative, in whole or in part, of a displayed object (in screen coordinate space) undergoes the traditional transformation and clipping functions in world coordinate space, and possibly mapping to screen coordinate space as described below. The first stage of processing data representative of a displayed primitive comprises transformation of the primitive in world coordinate space. Transformation may consist of a rotation, translation, scaling, shearing, or a combination of two or more of these functions, and is accomplished by matrix multiplication.

The second processing stage is clipping of the data against an operator defined clipping box (3D) or window (2D) in world coordinate space. Clipping discards that portion of a geometric primitive outside the box or window. The next stage is to map the clipped data to screen coordinate space, i.e., to a predefined viewport in the display screen. These processes are well known in the art and described in detail in the above cited Foley and Van Dam text. Because of its importance to describing the present invention, mapping will be examined in greater detail below.

Referring to Figures 3A and 3B, the predefined boundaries of a clipping window 34 in world coordinate space are:

xclip1,xclip2

x-clipping boundary

yclip1,yclip2

y-clipping boundary

zclip1,zclip2

z-clipping boundary (not shown).

(Although not described, the processing techniques disclosed herein are equally applicable to three dimensional clipping and the appended claims are intended to encompass such.) The coordinates for the corresponding operator defined viewport 32 in screen coordinate space are"

xview1,xview2

x-viewport boundary

yview1,yview2

y-viewport boundary

Thus, it will be observed that x- and y-coordinate window-­to-viewport ratios are expressed as:

wvx = (xview2 - xview1)/(xclip2 - xclip1)

wvy = (yview2 - yview1)/(yclip2 - yclip1)      (1)

Further, x- and y-coordinate window-to-viewport mapping equations are defined as:

xv = (xw - xclip1)*wvx + xview1

yv = (yw - yclip1)*wvy + yview1      (2)

In equations (2), xw and yw comprise the x- and y-coordinates of data representative of a primitive within clipping window 34 in world coordinate space (Fig. 3A), and xv and yv comprise the corresponding mapped x- and y-coordinates of the primitive within viewport 32 in screen coordinate space (Fig. 3B) from screen coordinate space to world coordinate space such that an inverse pick window 30 (Fig. 3A) is defined). The boundary of inverse pick window 36 is then used as the x-axis and y-axis clipping boundaries by processor 14 to initially identify in world coordinate space the operator selected geometric primi­tive(s) defining a displayed object, as will now be described with reference to Figure 4.

Upon entering the selection processing mode of the present invention at 60 "Start", for example in response to operator actuation of a function key, the processor is direct­ed to calculate the inverse pick window boundary 36 (Fig. 3A), "Calculate Inverse Pick Window Boundary" 62. The size of the pick window (in screen coordinates), psizex and psizey, is preconfigured by the operator. The center, pickx and picky, of the pick window in screen coordinate space is defined by the position of the operator controlled selector device 22. As shown in Figure 3B, pick or match window 30 in screen coordinate space is defined by four boundaries:

pwindx1 = pickx - psizex/2

pwindx2 = pickx + psizex/2

pwindy1 = picky - psizey/2

pwindy2 = picky + psizey/2      (3)

Applying inverse window-to-viewport mapping to pick window 30 results in inverse pick window 36 in world coordinate space with the following parameters:

x1 = (pwindx1 - xview1)*xratio + xclip1

x2 = (pwindx2 - xview1)*xratio + xclip1

y1 = (pwindy1 - yview1)*yratio + yclip1

y2 = (pwindy2 - yview1)*yratio + yclip1      (4)

wherein:

xratio = 1/wvx

yratio = 1/wvy

Window-to-Viewport ratios wvx and wvy are defined by equations (1) above. An inverse pick box (not shown) in 3D world coordinate space could alternatively be defined if desired, by using the preexisting z-axis clipping boundaries, i.e., zclip1 and zclip2, in combination with the inverse pick window parameters defined by equations (4).

After calculating the inverse pick window boundary, processor 14 is directed to replace the previously defined clipping window boundary 34 by the inverse pick window bound­ary 36, "Replace Clip Boundary By Inverse Pick Window Bound­ary" 64. Subsequently, the processor enters the main pick correlation processing loop at instruction 66 "If Not End Of Data, Fetch Data", where the processor is directed to obtain the next piece of data from the data list representative of displayed objects for performance of transformation and clipping, and possibly mapping and rasterization, operations thereon. Specifically, the processor transforms each piece of data "Transform Geometric Data" 68, and then clips it against the inverse pick window boundary 36, "Clip Against Inverse Pick Window Boundary" 70.

Preferably, trivial rejection clipping is utilized to further limit the processing time required for complex geo­metric primitives such as circles, ellipses, and filled area polygons. Curved objects, such as circles and ellipses, are typically represented by a plurality of line segments gene­rated directly from an appropriate curve parametric equation. These line segments are then individually transformer, clipp­ed, and mapped, a process which is very time consuming, and unnecessary if the transformed curved figure lies entirely outside the inverse pick window boundary. Thus, a trivial clip or extend test, which essentially generates a box around the primitive and determines whether the box intersects the clipping window, is preferably utilized to determine whether a primitive lies entirely outside the inverse pick window boundary. If outside, no additional processing is required and the processor returns via 71 to instruction 66, thereby avoiding mapping and rasterization of a nonselected primitive. In the case of filled polygons, the filing time is also saved where the primitive is trivially rejected as being outside the inverse pick window boundary. Trivial rejection processing is well known by those skilled in the art.

If intersecting inverse pick window 36, the clipped data is mapped to the viewport in screen coordinate space "Window-­To-Viewport Mapping" 72, and then rasterized, via pixel processor 16, "Rasterize Mapped Object" 74. The picking or selection of an object occurs during the rasterization step 74 when the generated pixels of an object are compared to the pick window. Some graphics processors have the ability to identify picked objects prior to mapping and rasterization. In such a device steps 72 and 74 may be omitted with process­ing proceeding directly from step 70 to step 76. Lastly, the processor stores the picked data "Store Picked Data" 76 and processes the data according to the rules set by the applica­tion program. Again, note that this processing technique eliminates the rasterization and examination of any geometric primitive not intersecting the inverse pick window, and therefore the pick window, thereby significantly reducing pick correlation time, especially when there are a number of complex primitives such as filled polygons or widened lines on the display screen.

As a further enhancement of the invention, an efficient zoom processing technique is provided. A "zoom" function is often available in graphics system to facilitate operator selection of a geometric primitive in an area of a complex display screen crowded with many displayed objects. As with the described pick window processing, only the contents of a zoom window are rasterized pursuant to this invention, thus significantly reducing processing time in comparison with prior zoom processing techniques, which again typically require rerasterization of the entire display screen.

Referring to Figure 5, when a operator selects the zoom mode, the processor begins at 80 "Start" and then proceeds to determine the zoom window boundary "Calculate Zoom Window Boundary" 82. As shown in Figure 6A, zoom window 110 is defined in world coordinate space as a rectangle which con­ tains an inverse pick window 112. As described before with reference to Figures 3A and 3B, the pick window 114 size in screen coordinate space is defined to be psizex and psizey, and its center, pickx and picky, in screen coordinate space is determined by the position of the operator controlled locator device. As before, applying inverse window-to-viewport mapping to pick window 114 (i.e. equations (3) and (4) above), values for parameters x1, x2, y1 and y2 (see Fig. 6A) defini­tive of inverse pick window 112 in world coordinate space are obtained.

Zoom window 110 is similarly determined by applying inverse window-to-viewport mapping (equations (4) above) to a rectangle in screen coordinate space (note shown) defined by:

pwtwox1 = pickx - 2psizex

pwtwox2 = pickx + 2psizex

pwtwoy1 = picky - 2psizey

pwtwoy2 = picky + 2psizey

Again, for purposes of explanation, the zoom window in world coordinate space is arbitrarily redefined herein to be sixteen times as large as the inverse pick window. The illustrated boundaries of zoom window 110 have values expressed by:

zoomwx1 = (pwtwox1 - xview1)*xratio + xclip1

zoomwx2 = (pwtwox2 - xview1)*yratio + xclip1

zoomwy1 = (pwtwoy1 - yview1)*yratio + yclip1

zoomwy2 = (pwtwoy2 - yview1)*yratio + yclip1      (6)

Wherein:

xratio = 1/wvx

yratio = 1/wvy

Window-to-Viewport mapping wvx, wvy, ratios are defined by equations (1). Xview1 and yview1 are the x-axis and y-axis minimum boundaries, respectively, of viewport 118 in screen coordinate space, and xclip1 and yclip1 are the x-axis and y-axis minimum boundaries, respectively, of the predefined clipping window in world coordinate space.

Continuing with Figure 5, after calculating the zoom window boundary, the processor is directed to replace the preexisting clipping boundary in world coordinate space with the zoom window boundary, "Replace Clip Boundary With Zoom Window Boundary" 84, and then to obtain the zoom viewport boundary, "Determine Zoom Viewport Boundary" 86. The zoom viewport 116 is a rectangular area in screen coordinate space of any size defined, i.e. preconfigured, by the operator. As shown in Figure 6B, viewport 116 is expressed by parameters zoomvx1, zoomvx2, zoomvy1, and zoomvy2. If desired, the zoom viewport can overlay the original data on the screen, similar to a pop-up window.

The processor next replaces the original viewport bound­ary 118 (Fig. 6B) with the zoom viewport boundary 116, "Re­place Viewport Boundary With Zoom Viewport Boundary" 88, and then calculates the new window-to-viewport ratio, "Recalculate Window-to-Viewport Ratio" 90. Recalculation of the window-to-­viewport ratio is necessary because, as noted, zoom viewport 116 is operator defined and possibly different in size from original viewport 118. Next, the processor enters the main processing loop at 92 "If Not End Of Data, Fetch Data", where processor 14 begins to work through the data list representa­tive of displayed geometric primitives. Again, each piece of data representative of a geometric primitive is transformed "Transform Geometric Data" 94, and clipped against the new clipping boundary, i.e., the zoom window boundary 110 (Fig. 6A), "Clip Against Zoom Window Boundary" 96. As before, those primitives outside the clipping boundary, i.e. the zoom window boundary, are discarded and the processor returns to instruc­tion 92 via line 97. Preferably, trivial clip testing is also utilized as described above to enhance processing time. Next, the processor is directed to map the retained clipped data to zoom viewport 116 in screen coordinate space using the recal­culated window-to-viewport ration, "Window-To-Viewport Mapp­ing" 98, and then to rasterize the mapped object, "Rasterize Mapped Object" 100. Thereafter, the processor stores the object for display on the screen within the zoom viewport, "Store Picked Data" 102. The operator can then use the locator device to initiate a select or pick operation from the geometric objects within the zoom viewport, which will be processed in the same manner as described with reference to Figure 4.

It will be observed from the above that this invention fully meets the objectives set forth herein. An interactive computer graphics display system processing method for identi­fying a displayed object which intersects an operator selected area of the display screen in a more efficient manner than heretofore known processing techniques is described. In addition, a zoom function processing method is described for enlarging an operator selected area of the display screen in a more efficient manner than heretofore known processing tech­niques is provided.

Although specific embodiments of the selection identi­fication processing method and the zoom processing method of the present invention have been illustrated in the accompany­ing drawings and described in the foregoing detailed descrip­tion, it will be understood that the invention is not limited to the particular embodiments described herein but is capable of numerous rearrangements, modifications and substitutions without departing from the scope of the invention. The following claims are intended to encompass all such modifica­tions.