Intra Prediction Coding Units for Video Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding systems face challenges in delivering high-quality video across a wide range of devices with different sizes, resolutions, and connectivity, as they require additional bandwidth and improved compression to accommodate varying device capabilities, which existing technologies have not adequately addressed.
Innovation Solution
A video coding system that generates a neighboring coding unit with a constrained intra prediction flag enabled, using only an intra-picture prediction mode, to create an intra block motion vector and a current coding unit, which is then used to generate a video bitstream for decoding and display on devices, enhancing compression efficiency and flexibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional video coding schemes are used to deliver high quality video across multiple devices, then video quality can be maintained, but additional bandwidth and improved compression are required
Solution Approach 1:
The video frame is divided into multiple coding units (current coding unit and neighboring coding units), with each unit processed independently using intra-picture prediction. This segmentation allows for more efficient local compression while maintaining overall video quality across different device resolutions.
Solution Approach 2:
The patent employs a constrained intra prediction flag that changes the prediction mode parameter, restricting reference samples to only those generated by intra-picture prediction rather than allowing inter-picture references. This parameter change improves compression efficiency by reducing temporal redundancy while adapting to different device capabilities.
2Quantity of substance
If video is compressed to reduce bandwidth requirements, then bandwidth consumption decreases, but video quality and features may be compromised
Solution Approach 1:
The neighboring coding units are generated in advance using intra-picture prediction before the current coding unit is processed. This preliminary action creates reliable reference samples that improve the accuracy of the current coding unit prediction, maintaining video quality while reducing the bitrate required.
Solution Approach 2:
The constrained intra prediction flag acts as an intermediary mechanism that controls which reference samples are available for prediction. By mediating between inter-picture and intra-picture references, it optimizes the balance between compression efficiency and video quality for different device requirements.
3Productivity
If intra-picture prediction mode is used for all coding units, then compression efficiency improves, but flexibility in handling different prediction scenarios is reduced
Solution Approach 1:
The system dynamically adjusts the prediction mode on a per-coding-unit basis using the constrained intra prediction flag. This allows the encoder to switch between intra-picture prediction (when beneficial for compression) and other modes (when needed for quality), providing both efficiency and flexibility adaptively throughout the video sequence.
Solution Approach 2:
Different regions of the video frame (current and neighboring coding units) are treated with different prediction strategies based on local characteristics. The constrained intra prediction flag enables local optimization where intra-picture prediction is applied to specific units that benefit from it, while other units can use alternative modes, achieving both efficiency and versatility.
Data Source
AI summary
A video coding system, and a method of operation thereof, includes: a source input module for receiving a frame from a video source; and a picture process module, coupled to the source input module, for generating a neighboring coding unit in the frame, for generating an intra block motion vector with a constrained intra prediction flag enabled and the neighboring coding unit generated based on only an intra-picture prediction mode, for generating a current coding unit based on the intra block motion vector, the current coding unit adjacent the neighboring coding unit, and for generating a video bitstream based on the current coding unit for a video decoder to receive and decode for displaying on a device.