Inter Prediction Merge Candidates for Better Motion Vector Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image encoding/decoding methods require performance improvements, particularly in image encoding and decoding processes.
Innovation Solution
An image encoding/decoding apparatus that modifies a motion vector predictor using an adjustment offset, constructing a motion information prediction candidate list, selecting a prediction candidate index, and deriving a prediction motion vector adjustment offset based on offset information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If existing image encoding/decoding methods are used, then the basic encoding/decoding function is maintained, but coding performance is insufficient
Solution Approach 1:
The motion vector prediction process is segmented into multiple stages: constructing candidate lists from spatial and temporal neighbors, selecting prediction candidates, and applying adjustment offsets. This segmentation allows each stage to be optimized independently, improving overall coding performance without overwhelming complexity
Solution Approach 2:
The patent performs preliminary actions by pre-construcing motion candidate lists and pre-calculating adjustment offsets before final motion vector determination. This preliminary preparation enables faster and more accurate motion compensation during decoding, enhancing coding performance while managing complexity through structured preprocessing
2Measurement precision
If motion vector prediction is performed without adjustment offsets, then the processing is simpler, but the prediction accuracy is lower
Solution Approach 1:
Adjustment offsets serve as intermediaries between the basic motion vector prediction and the final motion vector. These offsets compensate for prediction errors by incorporating information from neighboring blocks and temporal references, thereby improving prediction accuracy without requiring complete redesign of the prediction mechanism
Solution Approach 2:
The patent modifies the motion vector prediction by changing parameters such as the adjustment offset values derived from neighboring block motion vectors and temporal motion vectors. These parameter adjustments fine-tune the prediction accuracy, allowing the system to achieve better precision while maintaining a manageable level of computational complexity through controlled parameter modification
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An image decoding method performing inter prediction is presented. The method comprising: constructing a merge candidate list of a target block including a statistical merge candidate from a statistical merge candidate group and a spatial merge candidate from a neighboring block of the target block; determining a motion vector from a selected merge candidate in the merge candidate list based on a candidate index; and determining a prediction block for the target block based on the determined motion vector. The statistical merge candidate group includes one or more statistical merge candidates and a number of statistical merge candidates in the statistical merge candidate group cannot be larger than a predetermined maximum number of the statistical merge candidates. The statistical merge candidate indicates inter-prediction information of a block that is decoded prior to the target block and the merge candidate list includes one or more statistical merge candidates in the statistical merge candidate group that are not immediately adjacent to the target block. The statistical merge candidate group is updated after completion of decoding the target block by the following method: when the inter-prediction information of the target block is not overlapped with any of the statistical merge candidates in the statistical merge candidate group, the statistical merge candidate group is updated by adding inter-prediction information of the target block as the first statistical merge candidate of the statistical merge candidate group and removing the last statistical merge candidate from the statistical merge candidate group and modifying the orders of the remained statistical merge candidates; when the inter-prediction information of the target block is overlapped with any of the statistical merge candidates in the statistical merge candidate group, the statistical merge candidate group is updated by adding inter-prediction information of the target block as the first statistical merge candidate of the statistical merge candidate group and removing the overlapped statistical merge candidate from the statistical merge candidate group and modifying the orders of the remained statistical merge candidates. The statistical merge candidate group is initialized when the target block is the first block of a predetermined data unit. In the merge candidate list, the statistical merge candidate is included after the spatial merge candidate in order.