Temporal Trilinear Filtering for Video Rendering

The patent applies partial action by performing bilinear filtering instead of full trilinear filtering. It selectively applies filtering operations only when necessary (when scale changes are detected) and only at the current mipmap level, rather than always performing the complete trilinear filtering across multiple mipmap levels. This reduces the number of filter computations while still addressing aliasing artifacts in critical scenarios.

The patent changes the filtering parameter by switching between different mipmap levels based on detected scale changes. Instead of consistently applying trilinear filtering, it dynamically adjusts which mipmap level to use, changing the filtering intensity and type based on the specific rendering conditions. This allows the system to maintain image quality when needed while reducing computational overhead when scale changes are minimal.

The patent introduces dynamic adaptation by detecting scale changes between frames and dynamically selecting appropriate mipmap levels. The system monitors the scale of rendered pixels relative to the digital display and adjusts the filtering approach in real-time. When scale changes are detected, it switches to using different mipmap levels; when stable, it maintains the current level, avoiding unnecessary computational complexity.

The patent segments the filtering process by separating it into distinct stages: scale change detection, mipmap level selection, and filtering application. This segmentation allows the system to handle only the necessary portions of the filtering pipeline based on current rendering conditions, reducing overall complexity while maintaining quality where needed.

The patent extracts and removes the frame buffering requirement from the temporal anti-aliasing process. Instead of relying on buffered previous frames like TXAA, it uses real-time scale change detection between the current frame and the digital display to determine filtering needs. This extraction eliminates the memory overhead and processing complexity associated with maintaining and managing frame buffers while still achieving temporal anti-aliasing effects.

The system performs self-service by using its own scale change detection capability to determine when filtering is needed, rather than relying on externally buffered frame data. The scale detection mechanism inherently provides the temporal comparison information needed for anti-aliasing decisions, making the separate frame buffering infrastructure unnecessary.