Signaling Reference Frame Distances for Interlaced Video
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video compression and decompression techniques, such as those in Windows Media Video versions 8 and 9, and international standards like H.261, H.262, H.263, and H.264, do not effectively account for polarity changes and distance variations between current and reference fields in interlaced video frames, leading to inefficiencies in motion vector prediction for separately coded fields.
Innovation Solution
The implementation of techniques and tools for encoding and decoding reference frame distances, which involve signaling a count of frames between a current frame and a preceding reference frame, or using an arbitrary value to improve motion vector prediction accuracy, allowing for better handling of interlaced video frames by a video encoder or decoder.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If motion vector prediction is performed without accounting for reference frame distance and polarity changes, then device complexity is reduced, but motion vector prediction accuracy deteriorates
Solution Approach 1:
The patent applies parameter changes by introducing reference frame distance (REFDIST) as a new parameter that modifies the motion vector prediction process. The encoder signals REFDIST values that indicate the number of frames between the current frame and reference frames, and the decoder uses these parameters to adjust motion vector scaling and selection. This allows the prediction accuracy to be improved by adapting to varying frame distances and polarity changes without requiring complex additional hardware, as the changes are implemented through parameter adjustments in the existing prediction algorithms.
2Measurement precision
If reference frame distance information is signaled for each frame, then motion vector prediction accuracy is improved, but bit rate increases
Solution Approach 1:
The patent applies local quality by selectively signaling reference frame distance information only where it is most beneficial. Rather than encoding REFDIST for every frame unconditionally, the system signals this information locally at frames or segments where reference frame distance variations occur or where motion vector prediction would benefit most from this information. This localized approach provides accurate prediction where needed while minimizing the overall bit rate increase.
Solution Approach 2:
The patent implements partial action by signaling reference frame distance information for only some frames rather than all frames. The encoder determines on a frame-by-frame basis whether to signal REFDIST based on whether the current frame uses reference frames at different distances or polarities. This partial signaling approach provides sufficient prediction accuracy improvement while avoiding the excessive bit rate cost of universal signaling.
3Reliability
If conventional motion vector prediction is used without reference frame distance accounting, then processing speed is maintained, but video quality deteriorates
Solution Approach 1:
The patent applies feedback by using the signaled reference frame distance information to adjust the motion vector prediction process dynamically. The decoder receives REFDIST values and uses this feedback to scale and select motion vectors appropriately for the current frame's distance from reference frames. This feedback mechanism improves video quality by adapting prediction to actual frame distances while maintaining processing speed, as the adjustments are computed efficiently using the provided parameters rather than requiring complex reanalysis.
Data Source
AI summary
Techniques and tools for signaling reference frame distances are described. For example, a video encoder signals a code for a reference frame distance for a current field-coded interlaced video frame. The code indicates a count of frames (e.g., bi-directionally predicted frames) between the current frame and a preceding reference frame. The code may be a variable length code signaled in the frame header for the current frame. The encoder may selectively signal the use of a default value for reference frame distances rather than signal a reference frame distance per frame. A video decoder performs corresponding parsing and decoding.