Image Prediction Forward Motion Compensation Uncovered Zones
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing forward motion compensation techniques in video coding result in overlapping and uncovered zones due to different displacements on blocks, limiting the performance of the coding scheme and requiring complex restoration methods.
Innovation Solution
A method for predicting images using forward motion compensation that involves determining displacement information between reference blocks and corresponding blocks, forming enlarged blocks to include adjacent pixels, and calculating prediction values from these enlarged blocks, which can reconstruct uncovered points and adapt to displacements, thereby reducing complexity and improving coverage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If forward motion compensation is used to predict images from reference images, then prediction coverage is improved, but uncovered zones appear between blocks due to different displacements
Solution Approach 1:
The patent merges multiple reference blocks into an enlarged reference block that encompasses both the original reference block and adjacent blocks. This combination allows the prediction to cover areas that would otherwise remain uncovered, eliminating gaps between projected blocks while maintaining continuous coverage across the entire image.
Solution Approach 2:
The patent performs preliminary determination of displacement information for the enlarged reference block before actual prediction. By pre-calculating the displacement that represents the entire enlarged block (including adjacent areas), the method ensures that all regions are covered in advance, preventing uncovered zones from forming during the prediction process.
2Measurement precision
If template matching technique is used to restore uncovered zones, then restoration quality is improved, but implementation complexity increases significantly
Solution Approach 1:
The patent extracts only the essential displacement information from the enlarged reference block, rather than performing comprehensive template matching across multiple candidates. By taking out just the necessary displacement data needed for prediction, the method achieves adequate restoration quality without the computational burden of full template matching implementation.
Solution Approach 2:
The patent uses a simplified, computationally inexpensive displacement determination approach instead of complex template matching. This disposable-like simplified method provides sufficient restoration quality for video coding purposes while avoiding the high computational cost, making it suitable for real-time applications.
3Productivity
If conventional forward motion compensation is used, then processing speed is maintained, but uncovered zones limit coding performance
Solution Approach 1:
The patent dynamically adjusts the reference block size by creating an enlarged reference block that adapts to the specific displacement characteristics of the current prediction scenario. This dynamic enlargement ensures complete coverage for each prediction operation, eliminating uncovered zones and improving coding performance without requiring complex multi-pass processing that would slow down execution.
Data Source
Figure 1A~1B
Figure 2A~2B
Figure 3A~5
AI summary
The invention relates to a method for predicting an image by compensation during forward movement from at least one reference image. According to the invention, for at least one block of a reference image, referred to as a reference block, the prediction method (22) includes the following steps: determining (221) at least one piece of movement information, representing a movement between said reference block and a matching block in the image to be predicted; determining (222) at least one enlarged block, including said reference block and at least one point of a block adjacent to said reference block in said reference image; determining (223) a prediction value for at least one point of the image to be predicted from said at least one enlarged block and from said movement information determined for said reference block.