Image Processing Apparatus Quadtree Block Offset Calculation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In H.265/HEVC encoding, the complexity of calculating the position of each CU within a CTU due to varying block sizes complicates the encoding process, as offset amounts are not regular and require frequent recalculations based on reading order and size.
Innovation Solution
An image processing apparatus and method that specifies a second block to be encoded after a first block based on the size of the first block, using a quadtree spatial splitting approach, with a control unit initializing variables and employing a coordinate conversion table to determine the offset for encoding, thereby simplifying the encoding process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If recursive quadtree spatial splitting is used to divide CTU into CUs of varying sizes, then encoding efficiency and image quality are improved, but the complexity of calculating CU positions increases due to irregular offset amounts
Solution Approach 1:
The patent segments the CTU encoding process into distinct phases: first dividing the CTU into variable-sized CUs using quadtree spatial splitting, then separately calculating offset amounts for each CU based on its specific size and position. This segmentation allows the system to handle different CU sizes systematically without requiring complex real-time calculations during encoding.
Solution Approach 2:
The patent performs preliminary calculations of offset amounts for all CUs before the actual encoding process. By pre-calculating the offset amounts based on the quadtree splitting structure and storing them in a data structure, the system eliminates the need for complex runtime calculations, thereby reducing processing complexity while maintaining encoding efficiency.
2Measurement precision
If offset amounts are calculated based on reading order and CU size for each CU, then accurate positioning is achieved, but processing time increases due to frequent recalculations
Solution Approach 1:
The patent calculates and stores offset amounts for all CUs in advance, before the actual encoding process begins. By pre-computing these values based on the quadtree splitting hierarchy and Morton order reading sequence, the system achieves accurate positioning without requiring frequent recalculations during encoding, thus reducing processing time while maintaining positioning precision.
Solution Approach 2:
The patent creates a data structure that copies and stores the pre-calculated offset amounts for reference during encoding. Instead of recalculating offsets repeatedly, the system references this pre-computed data structure, which maintains positioning accuracy while significantly reducing the computational time required during the actual encoding process.
3Adaptability or versatility
If variable block sizes are used for CUs, then encoding flexibility is improved, but the regularity of offset patterns is lost requiring complex position specification
Solution Approach 1:
The patent applies local quality by calculating offset amounts specific to each CU's local characteristics (size and position within the CTU). Rather than using a uniform offset pattern for all CUs, the system computes customized offsets for each CU based on its specific attributes in the quadtree hierarchy, thereby maintaining encoding flexibility while managing position specification complexity through a systematic local adaptation approach.
Solution Approach 2:
The patent changes the offset parameter dynamically based on CU size and position within the quadtree structure. By adjusting the offset calculation parameters according to the specific CU being processed (8x8, 16x16, or 32x32 sizes), the system maintains flexibility for variable block sizes while reducing specification complexity through consistent parameter adjustment rules based on the quadtree splitting hierarchy.
Data Source
AI summary
A second block to be encoded after a first block, from among a plurality of blocks obtained by recursively dividing an image region in accordance with quadtree spatial splitting, is specified based on a size of the first block. The second block is encoded.