Hash Table Construction for Variable-Size Block Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video encoding technologies face challenges in efficiently performing motion estimation and block vector estimation due to high computational complexity and memory requirements, especially when dealing with variable-sized blocks, which hinders real-time encoding scenarios.
Innovation Solution
The implementation of hash table construction and hash-based block matching techniques, where hash values are determined for base-size candidate blocks in a reference picture, allowing for efficient identification of matching blocks in the reference picture, reducing computational and memory costs without compromising compression efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If hash values are stored for all available block sizes in the hash table, then block matching accuracy for variable-size blocks is improved, but computational cost and memory cost increase significantly
Solution Approach 1:
The patent segments the hash table construction by storing hash values only for base-size candidate blocks (e.g., 64x64 blocks) rather than all possible block sizes. This segmentation reduces the number of hash entries while maintaining matching accuracy through hierarchical block matching, where larger blocks are matched by comparing their constituent base-size blocks.
Solution Approach 2:
The patent introduces a hierarchical dimension to block matching by operating at multiple block size levels. Instead of storing hash values for every possible block size (2D approach), it uses a layered approach where base-size blocks form the foundation and larger blocks are constructed hierarchically, reducing the dimensional complexity of the hash table.
2Measurement precision
If traditional motion estimation and block vector estimation methods are used, then encoding accuracy is maintained, but encoding speed decreases due to computational intensity
Solution Approach 1:
The patent performs preliminary hash table construction using only base-size blocks before actual block matching. This preliminary action pre-organizes the reference data structure, enabling faster query operations during encoding without compromising the accuracy of motion estimation and block vector estimation.
Solution Approach 2:
The patent introduces hash-based block matching as an intermediary step between traditional motion estimation and final encoding. This intermediary uses hash values for rapid candidate identification, filtering out non-matching blocks before detailed comparison, thus maintaining accuracy while significantly improving encoding speed.
3Adaptability or versatility
If variable-size blocks are supported during encoding, then compression efficiency is improved, but memory requirements increase
Solution Approach 1:
The patent creates a universal hash table structure based on base-size blocks that serves multiple block size requirements. The same base-size hash table can support matching for various block sizes through hierarchical construction, eliminating the need for separate hash tables for each block size and reducing overall memory requirements while maintaining support for variable-size blocks.
Data Source
AI summary
Innovations in hash table construction and hash-based block matching for image encoding or video encoding are described. For example, an encoder determines hash values for base-size candidate blocks in a reference picture. The encoder stores, in a hash table, the hash values for the base-size candidate blocks. The encoder encodes a trial-size current block in a current picture. In some cases, the trial-size current block has a block size larger than the base block size. As part of the encoding, the encoder uses hash-based block matching, between base-size current blocks of the trial-size current block and the base-size candidate blocks, to identify a trial-size matching block, if any, in the reference picture. The encoder stores hash values only for the base-size candidate blocks. This can significantly reduce the computational cost and memory cost for hash table construction during encoding, without hurting compression efficiency or the overall speed of encoding.