Coding a vector
AssigneeIntel Corporation
InventorsAdam T. Lake, Christopher L. Gorman
Life of this patent
Estimating the term…
Abstract
An input vector is coded by obtaining a subdivision surface having vectors associated with bit patterns, and locating a target vector on the subdivision surface that is within a range of the input vector. The subdivision surface may be obtained by dividing an original surface. A bit pattern of the target vector is assigned to the input vector.
Claims20
A computer graphics system comprising:
- a graphics accelerator card for processing three-dimensional graphics data;
- a memory for storing executable instructions;
- a processor for executing the executable instructions to code the three-dimensional graphics data;
- and
- a network connection over which coded three-dimensional graphics data is output to a network;
- wherein the processor codes a selected vector in the three-dimensional graphics data using a three-dimensional subdivision surface having plural faces, each of the plural faces containing a single vector and having a numerical code, the processor executing instructions to perform coding by:
- receiving an input indicative of a level of coding accuracy;
- subdividing the subdivision surface in accordance with the input;
- identifying a face of the subdivision surface that contains a vector that is closest to being parallel to the selected vector;
- determining if the vector is within a predetermined range of the selected vector;
- wherein, if the vector is within the predetermined range, the processor assigns the selected vector a numerical code comprised of a numerical code of the face and of any faces previously subdivided to obtain the face;
- and
- wherein, if the vector is not within the predetermined range of the selected vector, the processor subdivides only the face, leaving other faces of the subdivision surface undivided, to produce a subdivided face and repeats identifying, determining, and assigning or subdividing using the subdivided face.
A method of coding an input vector using a three-dimensional subdivision surface having plural faces, each of the plural faces containing a single vector and having a numerical code, the method comprising:
- receiving an input indicative of a level of coding accuracy;
- subdividing the subdivision surface in accordance with the input;
- identifying a face of the subdivision surface that contains a vector that is closest to being parallel to the input vector;
- determining if the vector is within a predetermined range of the input vector;
- wherein, if the vector is within the predetermined range, the method further comprises assigning the input vector a numerical code comprised of a numerical code of the face and of any faces previously subdivided to obtain the face;
- and
- wherein, if the vector is not within the predetermined range of the input vector, the method further comprises subdividing only the face, leaving other faces of the subdivision surface undivided, to produce a subdivided face and repeating identifying, determining, and assigning or subdividing using the subdivided face.
An article comprising a machine-readable medium that stores executable instructions to implement a method of coding an input vector using a three-dimensional subdivision surface having plural faces, each of the plural faces containing a single vector and having a numerical code, the instructions causing a machine to:
- receive an input indicative of a level of coding accuracy;
- subdivide the subdivision surface in accordance with the input;
- identify a face of the subdivision surface that contains a vector that is closest to being parallel to the input vector;
- determine if the vector is within a predetermined range of the input vector;
- wherein, if the vector is within the predetermined range, the instructions cause the machine to assign the input vector a numerical code comprised of a numerical code of the face and of any faces previously subdivided to obtain the face;
- and
- wherein, if the vector is not within the predetermined range of the input vector, the instructions cause the machine to subdivide only the face, leaving other faces of the subdivision surface undivided, to produce a subdivided face and to repeat identifying, determining, and assigning or subdividing using the subdivided face.
An apparatus to implement a method of coding an input vector using a three-dimensional subdivision surface having plural faces, each of the plural faces containing a single vector and having a numerical code, the apparatus comprising circuitry to:
- receive an input indicative of a level of coding accuracy;
- subdivide the subdivision surface in accordance with the input;
- identify a face of the subdivision surface that contains a vector that is closest to being parallel to the input vector;
- determine if the vector is within a predetermined range of the input vector;
- wherein, if the vector is within the predetermined range, the circuitry assigns the input vector a numerical code comprised of a numerical code of the face and of any faces previously subdivided to obtain the face;
- and
- wherein, if the vector is not within the predetermined range of the input vector, the circuitry subdivides only the face, leaving other faces of the subdivision surface undivided, to produce a subdivided face and repeats identifying, determining, and assigning or subdividing using the subdivided face.
Dependencies are inferred from the claim text.