Dynamic depth-of-field emulation based on eye-tracking

A method for processing graphical data, the method comprising:

• receiving a stream of graphics primitives;
• generating a plurality of sample positions in a two-dimensional field, and
• computing depth values and color values for sample positions interior to the graphics primitives;
• assigning blur values to the interior sample positions based on the depth values and a concentration depth of a viewer, and storing the blur values and the color values for the interior samples of the graphics primitive in a sample buffer;
• subsequently reading the blur values and the color values corresponding to sample positions in a neighborhood of a current filter center from the sample buffer;
• assigning a filter coefficient to each sample position of the neighborhood based on a positional measure of the corresponding sample position with respect to the filter center and the corresponding blur value;
• multiplying each color value of the neighborhood by the assigned filter coefficient to generate a weighted color value, and computing a pixel value based on the weighed color values;
• providing a video signal for display on a display device, wherein at least a portion of the video signal is determined by the pixel value.

A computer system comprising:

• a display device;
• a host computer configured to execute a graphics application program;
• a graphics system coupled to the display device and the host computer, wherein the graphics system is configured to receive a stream of graphics primitives from the host computer in response to execution of the graphics application program, wherein the graphics system comprises:
• a rendering engine configured to sense a concentration depth of a viewer, to generate a plurality of samples in response to the graphics primitives, to assign a blur value to each of said samples based on the depth of said samples relative to the concentration depth;
• a sample buffer configured to store the samples and the blur value for each sample;
• a filtering engine configured to read a plurality of said samples and the corresponding blur values from the sample buffer, to filter said plurality of samples to generate a pixel value, and to transmit the pixel value to the display device, wherein said filtering operates on each of said samples with a corresponding filter function having a spatial cutoff frequency determined by the corresponding blur value.

A graphics system comprising:

• a rendering engine configured to generate depth values and sample color vectors for a plurality of sample positions, and assign chromatic distortion values to the sample positions based on data comprising the depth values and a concentration depth of a viewer, wherein the sample color vectors include a first color component and second color component for each sample position;
• a sample buffer configured to store the chromatic distortion values and the sample color vectors for the sample positions;
• a filtering engine configured to read the chromatic distortion values and the sample color vectors for the sample positions from the sample buffer, compute a first pixel color for an output pixel by filtering the first color components of the sample color vectors in a first neighborhood of a first position, compute a second pixel color for the output pixel by filtering the second color components of the sample color vectors in a second neighborhood of a second position, wherein a distance of separation between the first position and the second position is controlled by the chromatic distortion value;
• wherein the first pixel color and the second pixel color are usable to generate at least a portion of a displayable image.

A graphics system comprising:

• a rendering engine configured to generate depth values and distortion values for a plurality of samples in response to measurements of at least an instantaneous concentration depth of a viewer;
• a sample buffer configured to store the depth values and distortion values for the plurality of samples;
• and
• a filtering engine configured to generate pixel color components for an output pixel by filtering corresponding color components of the samples, wherein the distortion values of the samples determine an amount of distortion the filtering engine applies to the sample color components of the samples in a filtration process to generate the pixel color components;
• wherein the output pixel determines at least a portion of a displayable image.