Intra-Block Copy Prediction for Screen Content Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards, such as H.264/MPEG-4 AVC, face inefficiencies in compressing screen content, which is poorly handled due to its discontinuous tone nature, leading to suboptimal coding efficiency and increased bitstream size, while also requiring complex algorithms that hinder real-time decoding.
Innovation Solution
A method for copying blocks of video samples in a video bitstream using adaptation parameters to decode and identify the spatial location of blocks, allowing for efficient encoding and decoding of block vectors, particularly in HEVC, by reading and decoding multi-dimensional vectors from the bitstream to facilitate intra-block copy prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If existing video coding standards (H.264/MPEG-4 AVC) are used to compress screen content, then general video compression is achieved, but coding efficiency is suboptimal and bitstream size increases
Solution Approach 1:
The patent divides the video content into distinct regions: screen content regions and non-screen content regions. Different coding methods are applied to each region, allowing optimized handling of screen content through intra-block copy prediction while using conventional inter-prediction for other content, thereby improving overall coding efficiency and reducing bitstream size.
Solution Approach 2:
The patent applies different prediction methods to different spatial locations within the video frame. Screen content regions utilize intra-block copy prediction with block vectors referencing other blocks within the same frame, while non-screen regions use traditional inter-prediction, achieving locally optimized compression performance.
2Loss of information
If complex algorithms are used to improve compression performance, then coding efficiency improves, but real-time decoding becomes hindered
Solution Approach 1:
The patent segments the prediction process into simple, distinct modes: intra-block copy prediction for screen content and conventional inter-prediction for other content. This segmentation avoids the need for complex hybrid algorithms while maintaining high coding efficiency for screen content through dedicated prediction mechanisms.
Solution Approach 2:
The patent uses intra-block copy prediction where blocks of screen content are copied from other locations within the same frame using block vectors. This copying mechanism is computationally simpler than traditional inter-prediction requiring motion estimation across multiple reference frames, reducing algorithmic complexity while improving compression for screen content.
Data Source
AI summary
A method of copying a block of samples of a video bitstream, is disclosed. A plurality of adaptation parameters is read from a local memory store. Each adaptation parameter corresponds to a component of a multi-dimensional vector in the video bitstream. Each of the components of the vector is decoded from the video bitstream using the corresponding adaptation parameter. The block of samples is copied from the video bitstream. The spatial location of the block of samples is identified using the decoded components of the vector.