Video Motion Vector Prediction for Lower-Bit Skip-Like Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video codecs encode high-resolution or high-quality video content using limited encoding methods based on macroblocks, leading to inefficiencies in data compression and bit representation.
Innovation Solution
A video encoding and decoding method that searches for various prediction motion vector candidates using a motion vector, reducing bit representation by employing a pre-set prediction mode different from skip and merge modes, and generating prediction motion vectors based on motion prediction information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If conventional video codecs use limited encoding methods based on macroblocks, then the encoding process is simple, but data compression efficiency deteriorates
Solution Approach 1:
The video data is divided into multiple types of blocks including skip blocks, merge blocks, and non-skipping/non-merge blocks. Each block type is processed using specialized encoding methods tailored to its characteristics, allowing for more efficient compression while maintaining encoding manageability through structured segmentation.
Solution Approach 2:
The encoding method dynamically adapts to different block types by selecting appropriate prediction modes and motion vector representations. The system transitions between different encoding strategies (skip mode, merge mode, conventional mode) based on the characteristics of each block, optimizing compression efficiency without excessive complexity.
2Ease of operation
If conventional video codecs use predetermined macroblock sizes, then the encoding process is straightforward, but compressibility deteriorates
Solution Approach 1:
The video stream is segmented into different block categories (skip, merge, non-skipping/non-merge) with distinct encoding characteristics. This segmentation allows each block type to be compressed using the most appropriate method, improving overall compressibility while keeping the encoding process organized and manageable.
Solution Approach 2:
The encoding approach changes parameters such as motion vector precision and prediction mode selection based on the block type. Skip blocks use simplified representation while non-skipping blocks use more detailed motion compensation, allowing the system to adapt compression parameters to match the actual content characteristics.
3Loss of substance
If video data is transformed into frequency region coefficients, then compression is easier, but transformation complexity increases
Solution Approach 1:
The patent extracts and processes motion information separately from the main video data transformation pipeline. By identifying skip and merge blocks early and handling them with simplified methods, the system removes the need for complex frequency transformation in certain blocks, reducing overall transformation complexity while maintaining compression efficiency.
Solution Approach 2:
The encoding method performs preliminary classification of blocks into skip, merge, and non-skipping/non-merge categories before applying full transformation processes. This preliminary action allows the system to apply simplified encoding to suitable blocks, reducing the amount of complex transformation work required while maintaining compression effectiveness.
Data Source
AI summary
Provided is a video decoding method performed by a video decoding apparatus, the video decoding method including: determining prediction mode information of a current block and an index indicating a prediction candidate, from a bitstream; determining a prediction candidate list according to the prediction mode information; when the prediction mode information of the current block indicates a pre-set prediction mode, determining a motion vector indicated by the index indicating the prediction candidate from the prediction candidate list, and determining a prediction motion vector of the current block based on at least one of pieces of motion prediction information related to the motion vector; and determining a motion vector of the current block based on the prediction motion vector, wherein the pre-set prediction mode is a prediction mode different from a skip mode and a merge mode.