Guided Offset Correction for Video Frame Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Digital video streams face significant challenges in reducing errors in reconstructed frames due to lossy encoding, which affects computing and communication resources, and existing post-reconstruction loop filters often operate blindly without access to both source and reconstructed frames.
Innovation Solution
The method involves classifying pixels in a reconstructed frame using a classification scheme, determining offset values for each class, applying these offsets to reduce error, and selecting classes based on error rate reductions, allowing for guided loop restoration with access to both source and reconstructed frames.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If lossy encoding is used to reduce data size, then the amount of data in video streams is reduced, but errors and artifacts are introduced in reconstructed frames
Solution Approach 1:
The patent segments the reconstructed frame into multiple classes based on pixel characteristics (e.g., brightness levels, texture patterns). By dividing the frame into homogeneous regions, the system can apply class-specific offset corrections that are tailored to each segment's error characteristics, thereby improving overall reconstruction accuracy without requiring full-frame processing of all pixels.
Solution Approach 2:
The patent applies different offset correction strategies to different classes of pixels based on their local characteristics. Each class receives a customized offset value derived from comparing reconstructed pixels with original source pixels, allowing the system to address local errors specific to each region rather than applying a uniform correction across the entire frame.
2Use of energy by moving object
If blind filtering is used without access to source frames, then computational resources are reduced, but error reduction effectiveness is limited
Solution Approach 1:
The patent performs preliminary comparison between reconstructed pixels and original source pixels during the encoding phase to compute class offset values. These offset values are then embedded in the bitstream and applied during decoding, allowing the decoder to achieve accurate error correction without requiring access to the original source frames, thus reducing computational resources at the decoding stage while maintaining high error reduction effectiveness.
Solution Approach 2:
The system uses feedback from the comparison between reconstructed and original pixels to generate class-specific offset corrections. The encoder computes these offsets based on the actual error patterns observed in each class, and this feedback information is transmitted to the decoder, enabling the decoder to apply precise corrections without needing to recompute the offsets or access the original source frames.
3Measurement precision
If class-specific offset correction is applied to all pixel classes, then reconstruction accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent applies offset correction selectively to only those pixel classes that exhibit significant systematic errors, rather than uniformly correcting all classes. By identifying and correcting only the classes with notable errors, the system achieves substantial improvement in reconstruction accuracy while avoiding the computational overhead of processing every pixel class, thus balancing accuracy improvement with complexity management.
Data Source
AI summary
Reducing error in a reconstructed frame is described. Pixels of the frame are classified into classes based on a classification scheme. Offset values for each class of at least some of the classes are determined, and a respective offset value for a class is applied to each pixel of the class, resulting in offset-adjusted pixels for the class. For the classes, a respective error rate reduction in using the respective offset value for a class as compared to omitting the respective offset value is determined, where the respective error rate reduction is based on the pixels of the class in the reconstructed frame, the offset-adjusted pixels of the class, and co-located source pixels in a source frame decoded to generate the reconstructed frame. A subset of classes is selected for reducing error in the reconstructed frame based on the error rate reductions.