Adaptive Block Transform Selection for Lower-Volume Image Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high resolution and quality images leads to increased data volume, resulting in higher transmission and storage costs, necessitating improved image encoding/decoding techniques.
Innovation Solution
The method involves determining horizontal and vertical transform types for image blocks based on their width and height, regardless of intra prediction modes, using implicit multi-transform selection to enhance encoding/decoding efficiency through adaptive transform type selection, shuffling, rearrangement, and flipping.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If high resolution and quality images are used, then image quality is improved, but data volume increases leading to higher transmission and storage costs
Solution Approach 1:
The patent applies parameter changes by dynamically selecting transform types (DST-7, DCT-2, Hadamard) based on intra prediction modes and block characteristics. This adaptive transformation optimizes the compression ratio by matching the transform type to the specific image content and prediction mode, thereby reducing data volume while maintaining high image quality.
2Productivity
If traditional transform methods are used, then encoding process is simple, but encoding efficiency is insufficient for high resolution images
Solution Approach 1:
The patent implements dynamics by making the transform type selection adaptive and dynamic rather than static. The encoder dynamically chooses between DST-7, DCT-2, and Hadamard transforms based on real-time analysis of intra prediction modes and block characteristics, allowing the encoding process to adapt to different image content and optimize efficiency for high resolution images.
Solution Approach 2:
The patent changes the transform type parameter based on intra prediction modes and block characteristics. By conditioning the transform selection on these parameters, the system achieves higher encoding efficiency without requiring complex manual configuration, as the parameters themselves drive the transform selection automatically.
3Productivity
If fixed transform type is used regardless of prediction mode, then encoding process is simple, but encoding efficiency is suboptimal
Solution Approach 1:
The patent applies parameter changes by making transform type selection dependent on intra prediction mode parameters. Different prediction modes (planar, angular, etc.) trigger different transform types, allowing the system to optimize encoding efficiency for each prediction scenario while maintaining a systematic and manageable complexity through parameter-based control.
Solution Approach 2:
The patent applies local quality by selecting different transform types for different local regions or blocks based on their specific prediction modes and characteristics. Instead of applying a uniform transform to the entire image, the system adapts the transform type to local block properties, improving overall encoding efficiency through localized optimization.
Data Source
AI summary
Disclosed herein is an image encoding/decoding method and apparatus based on transform. An image decoding method includes determining a horizontal transform type and a vertical transform type of a current block, deriving a residual block of the current block by performing inverse transform with respect to the current block based on the determined horizontal transform type and vertical transform type, and reconstructing the current block based on the residual block. The determining of the horizontal transform type and the vertical transform type is performed based on at least one of a width or a height of the current block regardless of an intra prediction mode of the current block, when the current block is in an intra sub-block partitions (ISP) mode.