Video Coding Intra Block Copy Reference Area Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies face inefficiencies in intra block copy mode due to constraints on the reference area, particularly when reconstructing blocks within the same picture, leading to processing delays and increased complexity.
Innovation Solution
The proposed solution involves constraining the reference area for intra block copy by excluding certain CTUs, such as the left CTU, and using padding samples to facilitate faster reconstruction without waiting for all reference samples to be available, thereby optimizing the block vector constraints and improving processing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If the reference area is constrained by excluding certain CTUs (e.g., left CTU), then the processing delay is reduced and reconstruction can proceed faster, but the available reference samples are reduced which may affect prediction accuracy
Solution Approach 1:
The patent applies preliminary action by pre-defining the constrained reference area and padding sample generation rules before the actual reconstruction process. The padding samples are prepared in advance based on the constrained reference area configuration, allowing the decoder to immediately use these pre-prepared samples without waiting for additional processing, thus reducing processing delay while maintaining prediction accuracy through the padding mechanism.
Solution Approach 2:
The patent introduces padding samples as an intermediary element to bridge the gap between constrained reference area limitations and the need for sufficient reference samples. These padding samples act as mediators that are generated based on the constrained reference area and used to supplement the actual reference samples, enabling faster reconstruction while maintaining adequate prediction information.
2Productivity
If intra block copy mode is implemented with constrained reference area, then processing efficiency is improved, but the complexity of determining and managing the constrained area increases
Solution Approach 1:
The patent applies segmentation by dividing the reference area into distinct regions: the constrained reference area (excluding certain CTUs) and the padding area. This segmentation allows the system to manage complexity through structured organization, where each region has a clear definition and purpose. The constrained reference area is segmented at the CTU level, making it easier to determine and manage while maintaining processing efficiency.
Solution Approach 2:
The patent utilizes parameter changes by configuring the constrained reference area through controllable parameters (e.g., indicating which CTUs are excluded). This allows the system to adjust the reference area constraints based on different coding scenarios without fundamentally changing the overall architecture, thus managing complexity through parameter-based configuration rather than structural complexity.
3Speed
If padding samples are used to facilitate faster reconstruction, then processing speed increases, but the accuracy of the reconstructed block may be compromised due to use of padded rather than actual reference samples
Solution Approach 1:
The patent applies local quality by differentiating between the constrained reference area (where actual reference samples are used) and the padding area (where padding samples are used). This local differentiation allows the system to maintain high accuracy in regions with available reference samples while using padding samples strategically in constrained regions, balancing reconstruction speed and accuracy through localized application of different sampling strategies.
Data Source
AI summary
Aspects of the disclosure provide methods and apparatuses for video encoding/decoding. In some examples, an apparatus for video decoding includes receiving circuitry and processing circuitry. According to an aspect of the disclosure, the processing circuitry decodes prediction information of a current block in a current coding tree unit (CTU) within a picture from a coded video bitstream. The prediction information is indicative of an intra block copy mode. The processing circuitry further determines a reference area in the picture for the current block based on signals in the coded video bitstream that indicate constraints of the reference area, and then the processing circuitry reconstructs at least a sample of the current block based on reference samples of the reference area in the picture.