Computer graphics processing and selective visual display system – Computer graphics processing – Graphic manipulation
Reexamination Certificate
1999-05-10
2002-04-09
Luu, Matthew (Department: 2672)
Computer graphics processing and selective visual display system
Computer graphics processing
Graphic manipulation
C345S619000, C345S630000, C345S634000, C345S637000
Reexamination Certificate
active
06369830
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates generally to computer-implemented display systems, and more particularly to a system and method of rendering translucent and complex-shaped overlapping layers in a display system.
2. Description of Background Art
Many existing display systems are capable of compositing two or more display elements to generate a final image. In such systems, display elements often include overlapping layers, such as for example in a windowing system for a graphical user interface wherein on-screen elements, such as windows, may be moved around and placed on top of one another.
Rendering and displaying an image having two or more overlapping layers presents certain problems, particularly in determining how to render that portion of the image where the layers overlap. When the overlapping layers are opaque, the graphics system need only determine which layer is on “top”, and display the relevant portion of that layer in the final image; portions of underlying layers that are obscured may be ignored. However, when overlapping layers are translucent, more complex processing may be called for, as some interaction among picture elements (pixels) in each overlapping layer may take place. Accordingly, some calculation may be required to overlay the image elements in order to derive a final image.
Compositing techniques for performing these calculations are known in the art. See, for example, T. Porter et al., in “Compositing Digital Images”, in
Proceedings of SIGGRAPH
'84, 1984, pp. 253-59. Generally, however, such techniques are directed toward compositing only two layers at a time. When more than two layers are to be composited, a number of separate operations must be performed in order to generate the final image. This is generally accomplished by compositing image elements in a bottom-up approach, successively combining each new layer with the results of the compositing operations performed for the layers below.
This step-by-step compositing approach has several disadvantages. If the image is constructed in the frame buffer, on-screen flicker may result as the system writes to the frame buffer several times in succession. Alternatively, the image may be constructed in an off-screen buffer, thus avoiding on-screen flicker; however, such a technique requires additional memory to be allocated for the buffer, and also requires additional memory reads and writes as the final image is transferred to the frame buffer.
In addition, step-by-step generation of the final image may result in poor performance due to the large number of arithmetic operations that must be performed. Writing date to a frame buffer is particularly slow on many computers; therefore, conventional systems which write several layers to the frame buffer in succession face a particularly severe performance penalty.
Finally, such a technique often results in unnecessary generation of some portions of image elements that may later be obscured by other image elements.
Conventionally, arbitrarily shaped windows and layers are accomplished by dividing a window into rectangular areas and/or blocking out some portions of a rectangle to make a rectangular layer appear to be arbitrarily shaped. Such techniques often result in additional processing time for areas of the windows that may have no effect on the final image.
What is needed is a system and method for rendering translucent layers, which avoids the above-referenced deficiencies and allows compositing of multiple layers without causing screen flicker and without requiring additional off-screen buffers. What is further needed is a rendering system and method that improves the performance of prior art systems when generating images having multiple overlapping translucent layers. What is further needed is a rendering system and method that avoids unnecessary generation of image element portions that will be obscured by other image elements in the final image.
SUMMARY OF THE INVENTION
In accordance with the present invention, there is provided a rendering system and method that facilitates efficient compositing of translucent and complex-shaped overlapping layers. An alpha channel is provided for each layer, in order to allow transparency characteristics to be specified on a pixel-by-pixel basis within the layer. Rendering is performed in front-to-back sequence, so that the invention is able to avoid unnecessary computations on layer portions that do not contribute to the final image. In addition, in one embodiment, compositing is made more efficient by first subdividing the image area into rectangles, wherein the set of overlapping layers (and their ordering) for all points within a given rectangle is constant. Compositing operations can then be performed on each rectangle separately, so as to avoid complex and time-consuming bounds-checking operations inside the innermost loop of the processing operation.
The present invention renders two or more overlapping layers, each having an alpha channel, into a destination bitmap. The destination bitmap may be the frame buffer of a video system, for example. For each pixel in the destination bitmap, the present invention processes the layers having a corresponding pixel in front-to-back sequence, accumulating color and alpha values. Once the accumulated alpha channel reaches full opacity, the present invention stops working on that pixel; any pixels lying beneath the stopping point would be completely obscured (since full opacity has been reached) and need not be processed.
The present invention thus avoids reading pixel data that corresponds to pixels that do not contribute to the final image. This economy in image processing results in improved performance of the graphics system. In addition, if the destination bitmap is the frame buffer, flickering effects are reduced because each pixel in the destination bitmap is written at most once. The present invention also enables easy implementation of visual effects such as fading in or fading out of windows, as well as shadows, glow effects, and the like.
Additional performance enhancement is achieved, in one embodiment, by tagging the topmost layer that has changed since the last update. If the front-to-back compositing operation stops before reaching the tagged layer, it is not necessary to write the result to the output bitmap (the frame buffer), since the output bitmap already contains the same data, from the last update. The present invention can thus avoid unnecessary write operations in such a situation, and thereby improve system performance, particularly when output is being written to the frame buffer.
In one embodiment, additional optimization is achieved by initially subdividing the destination bitmap into a number of tiles, each tile containing the same stack of layers to be rendered. By performing such subdivision, the present invention is able to make determinations as to which layers may contribute to the final image on a tile-by-tile basis rather than on a pixel-by-pixel basis. This simplifies the bounds-checking process, and minimizes the number of calculations that need be made in the innermost loop of the process, thus improving overall performance.
The present invention can be implemented, for example, in a windowing system to provide efficient rendering of overlapping translucent windows in an improved graphical user interface.
REFERENCES:
patent: 4780709 (1988-10-01), Randall
patent: 5428724 (1995-06-01), Silverbrook
patent: 5499327 (1996-03-01), Satoh
patent: 5594467 (1997-01-01), Marlton et al.
patent: 5638499 (1997-06-01), O'Conner et al.
patent: 5831628 (1998-11-01), Shimizu
patent: 5877762 (1999-03-01), Young
patent: 0 775 971 (1997-05-01), None
Blinn, James F., “A Ghost in a Snowstorm”,IEEE Computer Graphics and Applications, Jan./Feb. 1998, pp. 79-89.
Blinn, James F., “Three Wrongs Make a Right”,IEEE Computer Graphics and Applications, Nov. 1995, pp. 90-93.
Blinn, James F., “Composting, Part 2: Practice”,IEEE Computer Graphics and Applications, Nov. 1994, pp. 78-82.
Blinn, Jam
Brunner Ralph T.
Graffagnino Peter
Apple Computer Inc.
Chung Daniel J
Fenwick & West LLP
Luu Matthew
LandOfFree
Rendering translucent layers in a display system does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Rendering translucent layers in a display system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Rendering translucent layers in a display system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2879547