Transform Core Pair Selection Based on Block Location
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video encoding methods require high encoding complexity due to the need to try multiple transform core pairs for optimal rate-distortion optimization, leading to increased processing time and efficiency losses.
Innovation Solution
The method determines a first transform core pair based on the location information of a target reconstructed image block, reducing the need to try multiple pairs and simplifying the encoding process by using pre-determined correspondences between location information and transform core pairs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple transform core pairs are tried for rate-distortion optimization, then encoding performance is improved, but encoding complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-determining the correspondence between target reconstructed image block locations and transform core pairs. Instead of trying multiple transform core pairs during encoding, the optimal transform core pair is selected in advance based on the location of the target reconstructed image block, thereby reducing encoding complexity while maintaining encoding performance
Solution Approach 2:
The patent changes the parameter selection strategy from trying multiple transform core pairs to selecting based on location information. By changing how the transform core pair is determined (from exhaustive search to location-based selection), the patent reduces encoding complexity while preserving the ability to achieve good rate-distortion optimization
2Measurement precision
If multiple transform core pairs are tried for rate-distortion optimization, then optimal transform selection is improved, but processing time increases
Solution Approach 1:
The patent performs the transform core pair selection in advance by establishing a correspondence between target reconstructed image block locations and transform core pairs. This preliminary action eliminates the need to try multiple transform core pairs during actual encoding, significantly reducing processing time while maintaining optimal transform selection through the pre-established location-based correspondence
3Reliability
If multiple transform core pairs are tried for rate-distortion optimization, then transform optimization is improved, but encoding overheads increase
Solution Approach 1:
The patent reduces encoding overheads by performing transform core pair selection in advance based on location information. The correspondence between target reconstructed image block locations and transform core pairs is established beforehand, eliminating the need to signal multiple transform core pair candidates during encoding, thereby reducing encoding overheads while maintaining transform optimization
Data Source
AI summary
An image encoding and decoding method, includes: determining location information of a target reconstructed image block of a current to-be-encoded image block, where the target reconstructed image block is a reconstructed image block used to determine motion information of the current to-be-encoded image block; determining a first transform core pair based on the location information of the target reconstructed image block; and transforming a residual signal of the current to-be-encoded image block based on the first transform core pair, to obtain a transform coefficient.