Motion Field Estimation via Trajectory Derivation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding techniques struggle to accurately capture true motion activity in video signals, leading to inefficiencies in compression and decoding processes.
Innovation Solution
The method involves reconstructing reference frames from encoded frame data, determining motion trajectory information by concatenating motion vectors from these reference frames, and using this information to estimate a motion field for the encoded frame. This motion field estimate is then used to determine a co-located reference frame, which is employed for inter-prediction during the decoding process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional motion estimation techniques are used, then the encoding process is simple, but the prediction quality is insufficient and cannot accurately capture complex motion activities
Solution Approach 1:
The patent segments the motion estimation process into multiple stages: first reconstructing reference frames, then determining motion trajectories by concatenating motion vectors, estimating motion fields using optimization algorithms, and finally generating co-located reference frames. This segmentation allows complex motion activities to be captured through progressive refinement while managing computational complexity at each stage.
Solution Approach 2:
The patent performs preliminary actions by first reconstructing reference frames and determining motion trajectories before performing the actual motion field estimation. The optimization algorithm uses initial motion vectors as starting points, and co-located reference frames are prepared in advance to improve the accuracy of subsequent prediction processes.
2Measurement precision
If motion vectors from multiple reference frames are concatenated to improve motion tracking, then the motion field estimation accuracy improves, but the computational resources required increase
Solution Approach 1:
The patent applies partial action by selectively concatenating motion vectors from reference frames based on their relevance to the current block being processed. The optimization algorithm performs motion field estimation only for blocks that require improved accuracy, rather than uniformly processing all blocks, thus reducing overall computational energy while maintaining precision where needed.
Solution Approach 2:
The patent changes parameters by adjusting the optimization algorithm's search range and convergence criteria based on the complexity of the motion in different regions. For areas with simple motion, fewer iterations are performed, while complex motion regions receive more computational resources, dynamically balancing accuracy and energy consumption.
3Reliability
If a co-located reference frame is determined using motion field estimation, then the inter-prediction accuracy improves, but the decoding process becomes more complex
Solution Approach 1:
The patent creates a co-located reference frame by copying and warping pixels from existing reference frames using the estimated motion field. This copied reference frame is then used for prediction, improving accuracy while keeping the decoding process manageable through efficient pixel copying operations rather than complex real-time calculations.
Solution Approach 2:
The co-located reference frame acts as an intermediary between the original reference frames and the current frame being decoded. It mediates the prediction process by providing a temporally and spatially aligned reference that simplifies the matching process, reducing decoding complexity while maintaining high prediction accuracy.
Data Source
AI summary
A motion field estimate determined using motion vector information of two or more reference frames of a current/encoded frame is used to derive a motion vector for inter-prediction of the current/encoded frame. Motion trajectory information, including concatenated motion vectors and locations of the current/encoded frame at which those concatenated motion vectors point, is determined by concatenating motion vectors of the reference frames. A motion field estimate is determined using the motion trajectory information and, in some cases, by interpolating unavailable motion vectors using neighbors. The motion field estimate is used to determine a co-located reference frame for the current/encoded frame, and an inter-prediction process is performed for the current/encoded frame using a motion vector derived using the co-located reference frame. During decoding, the motion field estimate may be determined using motion vectors signaled within a bitstream and without additional side information, thereby improving prediction coding efficiency.