Segmental Prediction for Depth and Texture Data in 3D Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding methods face challenges in effectively handling sharp transitions in prediction blocks, leading to deteriorated prediction quality and coding performance due to the loss of high-frequency information and pixel value distortions.
Innovation Solution
A segmental prediction process is introduced, where pixels in the prediction block are classified into segments based on their values or gradients, and each segment is processed differently to form a modified prediction block, which is used for encoding or decoding the current block.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If normal inter-prediction is used for prediction blocks, then coding simplicity is maintained, but prediction quality deteriorates due to loss of high-frequency information and sharp transitions
Solution Approach 1:
The patent divides the prediction block into multiple segments based on gradient analysis. Each segment is identified by detecting sharp transitions and high-frequency regions, allowing differential processing where segments with sharp transitions receive enhanced prediction treatment while uniform segments use standard prediction, thus improving overall prediction quality without uniformly increasing complexity across the entire block.
Solution Approach 2:
The patent applies different prediction strategies to different regions within the prediction block based on local characteristics. Regions with sharp transitions and high-frequency information are identified through gradient analysis and processed with enhanced prediction methods, while regions with smooth variations use conventional prediction, optimizing prediction quality locally without unnecessarily complicating the entire coding process.
2Loss of information
If reconstructed pixels are used for prediction values, then coding efficiency is maintained, but pixel value distortions occur due to loss of high-frequency information
Solution Approach 1:
The patent performs gradient analysis and segment identification on the prediction block before final prediction value generation. By pre-identifying regions with sharp transitions and high-frequency content, the system can apply appropriate enhancement strategies to these specific regions while maintaining standard efficient coding for uniform regions, thus recovering high-frequency information where needed without compromising overall coding efficiency.
Solution Approach 2:
The patent modifies prediction parameters dynamically based on local gradient characteristics. For segments identified with sharp transitions, the system adjusts prediction parameters to preserve high-frequency information, while for uniform segments, standard parameters are maintained to ensure coding efficiency. This selective parameter adjustment recovers lost high-frequency information without uniformly reducing coding efficiency.
Data Source
AI summary
A method and apparatus for processing a prediction block and using the modified prediction block for predictive coding of a current block are disclosed. Embodiments according to the present invention receive a prediction block for the current block and classify pixels in the prediction block into two or more segments. Each segment of the prediction block is then processed depending on information derived from each segment of the prediction block to form a modified prediction segment. The modified prediction block consisting of modified prediction segments of the prediction block is used as a predictor for encoding or decoding the current block.