Video Decoder Motion Vector Prediction Using Skip Candidates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In image compression methods like MPEG and H.264, the encoding of motion vectors increases data overhead, reducing the image compression ratio, and existing motion vector prediction methods are not effective when neighboring blocks have different motion vectors.
Innovation Solution
An apparatus for decoding a motion vector of a current prediction unit using spatial and temporal motion vector candidates, where the temporal skip candidate has a reference picture index of 0 and a motion vector from a temporal skip candidate picture, and the spatial skip candidate's motion vector is scaled when reference pictures differ.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If motion vector information is encoded and inserted as it is, then motion estimation can be performed, but compression ratio decreases because overhead increases
Solution Approach 1:
The patent extracts only the essential motion vector information needed for decoding by using skip mode, where complete motion vector data is omitted and only necessary reference indices are transmitted. This reduces the quantity of transmitted data while maintaining sufficient motion estimation capability through temporal and spatial skip candidates.
Solution Approach 2:
Instead of transmitting complete motion vector information, the patent uses partial information (reference picture indices and motion vectors from skip candidates) to reconstruct the current block. This partial action approach reduces overhead while achieving acceptable prediction accuracy through the skip candidate mechanism.
2Device complexity
If median of neighboring motion vectors is used for prediction, then simple prediction is achieved, but prediction accuracy decreases when neighboring blocks have different motion vectors
Solution Approach 1:
The patent dynamically selects between temporal skip candidate and spatial skip candidate based on the specific characteristics of the current block and available reference data. This dynamic adaptation allows the system to choose the most appropriate prediction method (temporal or spatial skipping) to maximize prediction accuracy for different scenarios rather than using a fixed median approach.
Solution Approach 2:
The patent changes the prediction parameters by using different reference pictures (temporal reference vs. spatial reference) and different motion vector sources (temporal skip candidate picture vs. spatial skip candidate). This parameter variation enables the system to adapt to different motion patterns and achieve better prediction accuracy when neighboring blocks have different motion vectors.
3Ease of operation
If temporal skip candidate with reference picture index 0 is used, then decoding simplicity is improved, but applicability decreases when reference picture 0 is not available
Solution Approach 1:
The patent dynamically switches between temporal skip candidate (using reference picture index 0) and spatial skip candidate based on the availability of reference pictures and the current block's characteristics. This dynamic selection maintains decoding simplicity when temporal reference is available while adapting to situations where spatial reference must be used instead, thus preserving versatility.
Solution Approach 2:
The patent creates a universal decoding mechanism that can handle multiple scenarios: temporal skip when reference picture 0 is available, spatial skip when temporal reference is not available, and fallback to median prediction when neither skip candidate is available. This multi-functional approach ensures the decoder can operate correctly across different reference picture configurations.
Data Source
AI summary
Provided is an apparatus for decoding a moving picture. An entropy decoding unit restores a quantization coefficient sequence from a bitstream. An inverse quantization/transform unit generates a residual block. An inter prediction unit generates a prediction block of a current block based on motion vector information. When the prediction block is encoded in skip mode, motion information of the current block is restored using an available spatial or temporal skip candidate and the prediction block of the current block is generated using the motion information. The temporal skip candidate includes a reference picture index and a motion vector, the reference picture index of the temporal skip candidate is set to 0, and a motion vector of the temporal skip candidate is a motion vector of the temporal skip candidate in a temporal skip candidate picture. A scan pattern for inversely scanning the plurality of subsets is the same as a scan pattern for inversely scanning coefficients of each subset. The quantization step size is generated by adding a quantization step size predictor and a remaining quantization step size predictor.