Motion Vector Retrieval with Clipped Offset Search Ranges
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video compression technologies, such as HEVC, face inefficiencies in motion compensation due to using the same motion information for all samples in a CU, leading to increased memory read bandwidth and prediction complexity with the ATMVP technology.
Innovation Solution
A method to determine a target offset vector within specific value ranges in orthogonal directions, reducing memory read times by pre-reading motion vectors in a defined area of the target picture, using syntax elements to define thresholds and POC for accurate motion vector determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If ATMVP technology is used to improve motion compensation accuracy by determining collocated sub-blocks, then motion compensation accuracy is improved, but memory read bandwidth and prediction complexity increase due to frequent reads from the corresponding picture
Solution Approach 1:
The patent applies preliminary action by determining value ranges for offset vector components before the actual motion compensation process. By pre-defining the search area boundaries based on the current block's position and the offset vector ranges, the system prepares the memory read operations in advance, limiting reads to only necessary areas and reducing overall prediction complexity while maintaining accuracy
Solution Approach 2:
The patent implements local quality by making different parts of the picture undergo different memory read operations. Specifically, blocks at different positions have different value ranges applied to them, creating position-specific search areas. This localized approach ensures that memory reads are concentrated in relevant regions rather than uniformly across the entire picture, reducing overall complexity while preserving motion compensation accuracy
2Device complexity
If motion information is uniformly applied to all samples in a CU, then device complexity is reduced, but motion compensation accuracy deteriorates due to ignoring varying motion features
Solution Approach 1:
The patent applies segmentation by dividing the current picture block into multiple sub-blocks, each processed with its own motion compensation parameters. This segmentation allows different motion features to be captured in different regions while maintaining manageable processing complexity through systematic handling of each sub-block using the value range methodology
Solution Approach 2:
The patent implements local quality by applying different motion compensation approaches to different regions. By determining specific value ranges for offset vectors based on block position and characteristics, the system tailors the motion compensation process to local motion features, improving accuracy without uniformly increasing complexity across the entire picture
3Measurement precision
If the search area for collocated sub-blocks is expanded to improve motion vector accuracy, then motion vector accuracy is improved, but memory read times and bandwidth increase
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the search area parameters (value ranges for offset vector components) based on the current block's position and characteristics. By modifying these parameters adaptively rather than using a fixed large search area, the system achieves sufficient motion vector accuracy while significantly reducing the memory read time and bandwidth requirements
Solution Approach 2:
The patent implements preliminary action by calculating and establishing the value ranges for offset vector components before performing the actual motion compensation search. This preliminary determination of search boundaries allows the system to efficiently locate collocated sub-blocks without unnecessarily scanning larger areas, thereby reducing memory read times while maintaining accuracy
Data Source
AI summary
This application provides a motion vector obtaining method and apparatus. The method includes: determining a target offset vector of a block and identifier information of a target picture, wherein the block comprises at least one sub-block; determining a location of the sub-block; determining, as a target location coordinate value of a collocated sub-block, a location coordinate value obtained by performing a clipping operation on an initial location coordinate value in a range, wherein the initial location coordinate value is based on the location of the sub-block and the target offset vector; and obtaining a motion vector of the sub-block based on a motion vector corresponding to the target location coordinate value. Thus, a range of the target offset vector is limited, so that a quantity of memory read times can be reduced in a process of obtaining the motion vector of the collocated sub-block.