Image element depth buffering using two buffers

• US 5,471,567 A
• Filed: 08/08/1991
• Issued: 11/28/1995
• Est. Priority Date: 08/08/1991
• Status: Expired due to Term

• Current Assignee: Lockheed Martin Corporation (Martin Marietta Corporation)
• Original Assignee: Bolt Beranek & Newman, Inc. (Verizon Communications Inc.)
• Inventors: Kenworthy, Mark L., Soderberg, Brian T.
• Primary Examiner(s): Jankus, Almis R.
• Application Number:
  US07/741,999
• Time in Patent Office: 1,573 Days
• Field of Search: 395/118-122, 395/129-132, 395/126-128, 395/133-135
• US Class Current: 345/615
• CPC Class Codes: G06T 15/503 Blending, e.g. for anti-aliasing

Abstract

Depth buffered anti-aliasing in a real time image generation system utilizing two separate buffers, one for combining attributes of object pixel definitions which are of less than full coverage and another for storing the attributes of each new object pixel definition which is of full coverage and which is closer to the viewpoint than any attributes currently stored. If the depth value in the partial buffer is closer the viewpoint than that in the full buffer, a set of attributes is output which is a weighted mixture of those stored in the two buffers.

Claims

1. In a real time image generation system operating from a database in which objects can be defined with greater precision than the pixel resolution of the display and in which a weight value is provided with each object pixel definition which indicates the degree of pixel coverage and a depth value is provided which represents distance from viewpoint, the method of combining on a pixel basis contributions from successively presented different objects using first and second buffers;

said method comprising;

inputting a succession of object pixels defined by respective attributes including said weight value and said depth value;

storing in a first buffer the attributes of each new object pixel definition whose weight value represents full coverage and whose depth value represents a position closer to the viewpoint than attributes currently stored in said first buffer;

combining contributions weighted in accordance with respective weight values from;

(i) the attributes of each new object pixel definition whose weight value represents less than full coverage and whose depth value represents a position closer to the viewpoint than attributes currently stored in said first buffer, and

(ii) any attributes previously stored in said second buffer

if the combined weight of the combined contributions is less than full coverage, storing the combined contributions in said second buffer;

if the combined weight of the combined contributions equals full coverage and if the depth value represents a position closer the viewpoint than attributes stored in said first buffer, storing the combined attributes in said first buffer; and

outputting a pixel definition defined by a set of attributes which is a mixture, weighted in accordance with respective weight values, of the attributes in the two buffers if the depth value in the second buffer is closer the viewpoint than the depth value in the first buffer and is equal to those attributes stored in said first buffer if the depth value in the first buffer is closer the viewpoint than the depth value in the second buffer.

2. The method as set forth in claim 1 wherein the attributes of each new pixel definition include color and new and stored color attributes are combined in proportion to respective weight values.

3. The method as set forth in claim 2 wherein the depth value stored in said second buffer is the nearer to the viewpoint of the respective new object definition and any attributes currently stored in said second buffer.

4. The method as set forth in claim 2 wherein the weight values of successively presented pixel definitions of less than full coverage are cumulated in said second buffer.

5. In a real time image generation system operating from a database in which objects can be defined with greater precision than the pixel resolution of the display and with attributes for each pixel which include color and a weight value which indicates the degree of pixel coverage and a depth value is provided which represents distance from viewpoint, the method of combining on a pixel basis contributions from successively presented unordered different objects using first and second buffers;

said method comprising;

reading successive object pixel definitions;

storing in a first buffer the attributes of each new object pixel definition which is of full coverage and which is closer to the viewpoint than attributes currently stored in said first buffer;

combining weighted contributions from;

(i) the attributes of each new object pixel definition which is of less than full coverage and which is closer to the viewpoint than attributes currently stored in said first buffer, and

(ii) any attributes previously stored in said second buffer

the color attributes being combined in proportion to the respective weight values and the weights being cumulated;

if the combined weight of accumulated contributions is less than full coverage, storing the accumulated contributions in said second buffer;

if the combined weight of the accumulated contributions equals full coverage, storing the combined attributes in said first buffer; and

outputting a set of attributes which is a weighted mixture of the attributes in the two buffers if the depth value in the second buffer is closer the viewpoint than the depth value in the first buffer and is equal to those attributes stored in said first buffer if the depth value in the first buffer is closer the viewpoint than the depth value in the second buffer.

6. The method as set forth in claim 5 wherein the depth value stored in said second buffer is the nearer to the viewpoint of the respective new object definition and any attributes currently stored in said second buffer.

7. In a real time image generation system operating from a database in which objects can be defined with greater precision than the pixel resolution of the display and in which a weight value is provided with each object pixel definition which indicates the degree of pixel coverage and a depth value is provided which represents distance from viewpoint, the method of combining on a pixel basis contributions from successively presented different non-ordered objects using first and second buffers;

said method comprising;

as successive object pixel definitions are provided, storing in a first buffer the attributes of each new object pixel definition which is of full coverage and which is closer to the viewpoint than attributes currently stored in said first buffer;

combining weighted contributions from;

(i) the attributes of each new object pixel definition which is of less than full coverage and which is closer the viewpoint than attributes currently stored in said first buffer, and

(ii) any attributes currently stored in said second buffer

storing the accumulated contributions in said second buffer; and

outputting a set of attributes which is a weighted mixture of the attributes in the two buffers if the depth value in the second buffer is closer the viewpoint than the depth value in the first buffer.

8. In a real time image generation system operating from a database in which objects can be defined with greater precision than the pixel resolution of the display and in which a weight value is provided with each object pixel definition which indicates the degree of pixel coverage and a depth value is provided which represents distance from viewpoint, apparatus for combining on a pixel basis contributions from successively presented different objects;

said apparatus comprising;

first and second frame buffers;

means for storing in said first buffer, as successive object pixel definitions are read, the attributes of each new object pixel definition which is of full coverage and which is closer the viewpoint than attributes currently stored in said first buffer;

means for combining weighted contributions from;

(i) the attributes of each new object pixel definition which is of less than full coverage and which is closer the viewpoint than attributes currently stored in said first buffer, and

(ii) any attributes previously stored in said second buffer

means for storing in said second buffer the accumulated contributions if the combined weight of accumulated contributions is less than full coverage;

means for storing the combined attributes in said first buffer if the combined weight of the accumulated contributions equals full coverage and if the depth is closer the viewpoint than attributes stored in said first buffer; and

means for outputting a set of attributes which is a weighted mixture of the attributes in the two buffers if the depth value in the second buffer is closer the viewpoint than the depth value in the first buffer and is equal to those attributes stored in said first buffer if the depth value in the first buffer is closer the viewpoint than the depth value in the second buffer.