Image Decoding of Segmented Coefficient Strings for Variable Block Sizes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing image encoding methods, such as H.26L, face complexity and increased memory requirements due to the need for a large number of encoding tables and complicated procedures when applying Context-based Adaptive Variable Length Code (CAVLC) to Adaptive Block size Transforms (ABT) with varying block sizes, which complicates efficient entropy coding.
Innovation Solution
Divide the coefficient string from blocks of larger sizes into multiple coefficient strings of the same length as those from the smallest block size, allowing for entropy coding adapted to the minimum block size, thereby simplifying the encoding process and reducing memory requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If CAVLC is applied to ABT with varying block sizes, then encoding efficiency is improved, but device complexity and memory requirements increase due to the need for multiple encoding tables
Solution Approach 1:
The patent segments the coefficient string from larger blocks into multiple coefficient strings of the same length as those from the minimum block size. This allows the use of a single set of encoding tables designed for minimum block sizes, avoiding the need to create and manage multiple encoding tables for different block sizes, thus reducing device complexity while maintaining encoding efficiency.
Solution Approach 2:
The patent makes the encoding tables universal by designing them to work with coefficient strings of a fixed length corresponding to the minimum block size. By segmenting larger block coefficient strings into multiple fixed-length segments, the same encoding tables can be reused across all block sizes, eliminating the need for separate encoding tables for each block size and simplifying the encoding procedure.
2Productivity
If CAVLC is applied to ABT with varying block sizes, then encoding efficiency is improved, but memory requirements increase due to the need for multiple encoding tables
Solution Approach 1:
The patent segments the coefficient string from larger blocks into multiple coefficient strings of the same length as those from the minimum block size. This allows the use of a single set of encoding tables designed for minimum block sizes, avoiding the need to create and manage multiple encoding tables for different block sizes, thus reducing device complexity while maintaining encoding efficiency.
Solution Approach 2:
The patent makes the encoding tables universal by designing them to work with coefficient strings of a fixed length corresponding to the minimum block size. By segmenting larger block coefficient strings into multiple fixed-length segments, the same encoding tables can be reused across all block sizes, eliminating the need for separate encoding tables for each block size and simplifying the encoding procedure.
3Measurement precision
If multiple encoding tables are used for different block sizes, then entropy coding accuracy is improved, but ease of operation deteriorates due to complicated selection procedures
Solution Approach 1:
The patent segments the coefficient string from larger blocks into multiple coefficient strings of the same length as those from the minimum block size. This allows the use of a single set of encoding tables designed for minimum block sizes, avoiding the need to create and manage multiple encoding tables for different block sizes, thus reducing device complexity while maintaining encoding efficiency.
Solution Approach 2:
The patent makes the encoding tables universal by designing them to work with coefficient strings of a fixed length corresponding to the minimum block size. By segmenting larger block coefficient strings into multiple fixed-length segments, the same encoding tables can be reused across all block sizes, eliminating the need for separate encoding tables for each block size and simplifying the encoding procedure.
Data Source
AI summary
An image decoding method, apparatus and program for decoding coded picture data is disclosed. In the method, apparatus and program, the coded picture data is entropy-decoded to derive four transform coefficient strings each comprising entropy-decoded sixteen (16) transform coefficients. The four transform coefficient strings are combined into an 8x8 block of entropy-decoded sixty four (64) transform coefficients by de-interleaving the four strings in such a manner as to select a lower frequency transform coefficient alternately from the four strings and arrange it in the 8x8 block in an order of a low frequency to a high frequency. An inverse orthogonal transform is then performed on the 8x8 block of the entropy-decoded sixty four (64) transform coefficients.