Adobe, Inc.

S3 Graphics Co., Ltd.

Axel Schildan

Konstantine I Iourcha

Roger Swanson

Gavin Miller

Frank Jargstorff

Kevin Michael Goldsmith

Roger W Swanson

Various embodiments herein provide one or more systems, methods, software, and data structures for executing multiple functions using more than one resource within a computing system. Examples of functions comprise data processing filters and effects. Resources include, for example, a central processing unit (CPU) and a graphics processing unit (GPU). Based on an input graph received from a user, one or more execution graphs may be generated to indicate the resources to be used and data conversions to be performed.

System and method for executing multiple functions execution by generating multiple execution graphs using determined available resources, selecting one of the multiple execution graphs based on estimated cost and compiling the selected execution graph

Systems, methods, and apparatus, including software tangibly stored on a computer readable medium, involve image processing. An image processing graph includes multiple nodes that each correspond to an image processing operation. Input image data relating to a particular input image is received. Region of interest data defining an output image region to be generated based on the input image data is received. A domain of defined node output data is identified for each of the nodes using the input image data, and the graph is traversed in a first traversal order. An identification of the first traversal order is stored. A region of requested node output data is identified for each of the nodes using the region of interest data, and the graph is traversed in a second traversal order. The second traversal order is based at least in part on the stored identification of the first traversal order.

Region reasoning for image processing graphs

A system and method is provided for decreasing the amount of data required to represent depth information for 3D images. In accordance with one embodiment, depth information is represented by a piecewise function Z(x,y). An (x,y) space is split into areas representing a region of primitive shapes. For each of these regions, Z(x,y) is defined as a simple parametric analytical function. Subsequently, only a few parameters are required to encode this function in each region. By using these parametric analytical functions to represent depth value of the split, the present invention achieves advantages such as reductions in required storage space and required bandwidth with a concomitant increase in processing speed.

Area and span based Z-buffer

Growing community of inventors

Sunnyvale, CA, United States of America

Axel Schildan