VOXEL TRANSFER CIRCUIT FOR ACCELERATED VOLUME RENDERING OF A GRAPHICS IMAGE

Patent №

US 6,573,893

Granted

2003-06-03

Filed 2000

Owner

HEWLETT-PACKARD COMPANY

Lab

AI components

1

hardware

Assignment

Recorded

Dataset

AIPD

2023_r1 edition

Application

09704185

A voxel transfer circuit that implements high-speed voxel transfer operations. When implemented in a 2D or 3D texture mapping circuit of a graphics system, or in a graphics card adapter of a workstation, the voxel transfer circuit provides a comprehensive, accelerated volume rendering capability to the existing imaging pipeline. For each voxel in a volume data set, the voxel transfer circuit generates a texel having an opacity value and a color value based on one or more properties of the voxel including the local gradient magnitude and a lighting model. The relationship between these values is referred to herein as a classification or transfer function. In the present invention, the voxel transfer circuit implements a transfer function that employs a gradient-influenced classification. Such a transfer function generates opacity and color values of structures with excellent edge and surface discrimination. Preferably the voxel transfer circuit also includes the influence of a lighting model on the voxel, providing improved three-dimensional clarity and depth cues. As a result, the present invention enables existing systems to perform complete volume rendering at accelerated texture mapping speeds. Thus, the invention allows for modifications to the parameters used during the voxel transfer process, enabling the operator to interactively modify the results of the volume rendering operations at accelerated rates. This allows graphics applications to expand the use of volume rendering while also providing operators with the ability to interactively explore different ways of visualizing their 3D volume data. Thus, the present invention enables the operator to re-render the voxel repeatedly, gradually modifying the transfer function in small increments and view the results of the modification in a sequence of volume-rendered frames that are rendered quickly.

AI classification

AI hardware1.00
Planning0.20
Vision0.06
Evolutionary computation0.00
Machine learning0.00
Speech0.00
Knowledge representation0.00
Natural language0.00

Ownership

HEWLETT-PACKARD COMPANY

assignment · 116440255

Assignors

NAQVI, HASAN SHAHZAD, BURRELL, JEFF WILLIAM, LICHTENBELT, BARTHOLD

On an employer assignment, the assignors are typically the inventors.

© 2026 NYSGPT2525 LLC