Foveated Anti-Aliasing With Peripheral Frame Accumulation

The patent applies preliminary action by accumulating frame buffers from previous low-resolution renders before the current high-resolution render completes. The accumulation image is prepared in advance and then blended with the current render, allowing the system to maintain anti-aliasing benefits without requiring every single frame to be rendered at full resolution, thus reducing energy consumption while preventing flickering.

The patent applies local quality by differentiating the rendering resolution based on spatial location. High resolution is applied locally to the fovea region where users focus their attention, while low resolution is applied to the periphery regions. The accumulation blending is selectively applied only to peripheral regions where temporal aliasing occurs, preserving energy savings in the periphery while maintaining quality where needed.

The accumulation image is prepared in advance from previous frames and stored in a frame buffer. When generating the current output image, the system blends the pre-prepared accumulation image with the current render, eliminating the need to wait for previous frames to complete fully, thus maintaining high rendering speed while reducing flickering through the anti-aliasing blend operation.

The patent maintains continuity of useful action by continuously accumulating frame buffers from all previous renders. This continuous accumulation process ensures that the anti-aliasing blend always has access to historical data, smoothing out temporal variations and eliminating flickering while the system maintains high rendering throughput through the optimized foveated rendering pipeline.

The patent implements local quality by applying high resolution only to the fovea region and low resolution to the periphery region. The accumulation blending is applied selectively to peripheral regions where temporal aliasing occurs, while the fovea region maintains high resolution directly. This spatially adaptive approach improves overall image quality in critical regions while reducing computational complexity in less critical regions.

The patent segments the image into different regions based on resolution requirements. The fovea region is separated from the periphery region, with each region processed at different resolution levels. The accumulation frame buffer is also segmented, with blending applied only to peripheral regions. This segmentation allows the system to optimize computational resources by applying high computational complexity only where it is most needed for perceived image quality.