Foveated Image Compression in Portable Computing Devices

The patent applies different compression methodologies to different regions of the image frame based on their importance. The fixation sector (containing the user's focus point) is compressed with a first methodology that preserves quality, while foveated sectors (peripheral regions) use a second methodology that achieves higher compression. This local differentiation allows the system to maintain QoS for critical regions while reducing overall memory and bandwidth requirements through aggressive compression of less critical peripheral regions.

The patent implements region-specific compression where the fixation sector uses a compression methodology that maintains high quality suitable for user focus areas, while foveated sectors use more aggressive compression. This approach ensures that bus bandwidth is allocated efficiently - sufficient for transmitting quality data in the fixation sector while allowing higher compression ratios in peripheral sectors, thereby reducing overall bus bandwidth requirements while maintaining acceptable QoS.

The patent introduces dynamic, region-adaptive compression that adjusts compression intensity based on the specific characteristics of different image regions and their relative importance to user perception. By using fixation point information to identify critical sectors and applying appropriate compression methodologies to each sector, the system dynamically optimizes the balance between quality and compression ratio, eliminating the need to size components for worst-case scenarios across the entire image while maintaining QoS guarantees for perceptually important regions.

The patent applies compression factors locally based on region importance. The fixation sector, which contains the user's focus point and is therefore most critical to perceived quality, is compressed with a first methodology that maintains high output quality. Foveated sectors, which represent peripheral visual regions less critical to user perception, are compressed with a second methodology that achieves higher compression factors. This local quality approach allows aggressive compression in peripheral regions without significantly impacting overall perceived quality, thereby optimizing the balance between resource usage and output quality.