Video Block Decoding With Constrained Transform and CTU Sizes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards face challenges in efficiently decoding video frames due to the complexity and variability of video formats, particularly in handling different chroma formats and frame rates, which can lead to inefficiencies in processing and resource utilization.
Innovation Solution
Implementing a method for decoding video frames by constraining and signaling maximum transform and coding tree unit sizes using fixed-length codewords, allowing for flexible subprofiling and reducing unnecessary tool usage, thereby optimizing decoding processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If video coding standards support multiple chroma formats and variable frame rates, then versatility and adaptability improve, but decoding complexity and resource utilization worsen
Solution Approach 1:
The patent applies parameter changes by introducing constrained maximum transform block sizes and maximum coding tree unit sizes that can be signaled to the decoder. These parameters limit the range of transform and coding operations that must be supported, thereby reducing decoding complexity while maintaining compatibility with various video formats through selective parameter signaling.
Solution Approach 2:
The patent segments the video decoding process by dividing coding tree units into smaller transform blocks with constrained maximum sizes. This segmentation allows the decoder to process smaller, more manageable blocks rather than handling large variable-sized blocks, reducing computational complexity while preserving format versatility through flexible block size selection.
2Productivity
If video coding standards include a wide selection of compression tools, then compression performance improves, but implementation cost and device complexity worsen
Solution Approach 1:
The patent extracts and removes certain transform and coding tools from the decoder implementation by constraining their maximum block sizes. By signaling these constraints to the decoder, the encoder effectively removes the need for the decoder to implement full-range transform and coding operations, reducing implementation cost while maintaining high compression performance through selective tool usage.
Solution Approach 2:
The patent applies partial action by implementing only a subset of available transform and coding tools with constrained block sizes rather than supporting all possible tools and block sizes. This partial implementation achieves sufficient compression performance for the target application while significantly reducing implementation cost and device complexity.
3Productivity
If maximum transform and coding tree unit sizes are constrained and signaled, then decoding efficiency improves, but bitstream signaling overhead increases
Solution Approach 1:
The patent changes the parameter representation by using constrained maximum block size values that can be efficiently signaled. By limiting the range of possible block sizes, the patent enables more compact parameter signaling in the bitstream, reducing signaling overhead while improving decoding efficiency through predictable block size constraints.
Data Source
AI summary
Decoding an image frame from a bitstream, the image frame being divided into a plurality of coding tree units. The method comprises decoding a maximum transform block size constraint and or a maximum coding tree unit (CTU) size constraint from the bitstream; and decoding a maximum enabled transform block size and/or a maximum enabled CTU size from the bitstream. The decoded maximum enabled transform block size is less than or equal to the decoded maximum transform block size constraint. The decoded maximum enabled CTU size is less than or equal to the decoded maximum CTU size constraint. Determining each of the one or more transform blocks for each of the plurality of coding tree units according to the decoded maximum enabled transform block size, maximum enabled CTU and split flags decoded from the bitstream; and decode each of the determined one or more transform blocks from the bitstream.