Image Coding Tile Set Size Limiting for Parallel Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-resolution image decoding using multiple image decoding apparatuses leads to increased memory costs and power consumption due to the need for large horizontal line buffers and inter-chip communication for referencing pixels across different tile sets.
Innovation Solution
The image coding and decoding apparatuses divide the picture into tile sets with a maximum size of horizontal and vertical pixels limited to 2048, allowing independent decoding and reducing the need for inter-chip communication by ensuring all tiles are part of a Motion-Constrained Tile Set (MCTS) with restricted pixel sizes, thereby minimizing memory usage and power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If high-resolution pictures are decoded using multiple image decoding apparatuses, then decoding capability and parallel processing are improved, but memory costs and power consumption increase due to large horizontal line buffers and inter-chip communication requirements
Solution Approach 1:
The picture is divided into multiple tile sets, with each tile set further divided into tiles. This segmentation allows each decoding apparatus to independently decode specific tile sets without requiring large horizontal line buffers, thereby reducing memory costs and power consumption while maintaining parallel decoding capability
Solution Approach 2:
The patent extracts and removes the requirement for large horizontal line buffers by implementing independent tile-set-based decoding. Each decoding apparatus processes only its assigned tile sets with minimal buffer requirements, eliminating the need for extensive inter-chip communication and reducing overall system power consumption
2Productivity
If high-resolution pictures are decoded using multiple image decoding apparatuses, then decoding capability and parallel processing are improved, but memory costs increase due to large horizontal line buffers
Solution Approach 1:
By segmenting the picture into multiple tile sets that can be independently decoded, each decoding apparatus requires only small local buffers instead of large horizontal line buffers. This segmentation strategy significantly reduces memory requirements while preserving parallel decoding functionality
Solution Approach 2:
The patent removes the dependency on large horizontal line buffers by implementing a decoding architecture where each apparatus independently processes its assigned tile sets. This extraction of the buffer requirement eliminates the associated memory costs while maintaining high-resolution decoding capability
3Device complexity
If tile sets are designed with larger sizes to reduce the number of tile sets, then device complexity is reduced, but the ability to perform independent decoding and reduce inter-chip communication is compromised
Solution Approach 1:
The patent segments the picture into multiple small tile sets that can be independently decoded. This segmentation maintains ease of operation by allowing each decoding apparatus to work independently on its assigned tile sets, reducing inter-chip communication requirements while managing device complexity through standardized tile set structures
Data Source
AI summary
An image coding apparatus that codes a moving image composed of a plurality of pictures, includes a determination unit configured to define a tile set in a manner that a size of the tile set including one or more tiles obtained by dividing a picture, which corresponds to a size of the tile set at which coding can be independently performed, is set to be smaller than or equal to a predetermined size, and a coding unit configured to code the picture on the basis of the tile set defined by the determination unit.