H.264 Intra Prediction Datapath Redundant Calculation Elimination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for intra prediction in the H.264 digital video coding standard are inefficient due to redundant calculations and large chip real estate area costs, as they use a single datapath for all prediction modes, leading to low computational efficiency and increased bandwidth requirements.
Innovation Solution
A method is introduced that combines and performs calculations based on shared equations in parallel, using a two-level addition process for certain prediction modes, reducing redundant computations and improving efficiency without significantly increasing chip area, by selecting and processing neighboring pixel values according to specific prediction modes and block sizes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single datapath is used for all prediction modes, then device complexity is reduced, but computational efficiency deteriorates due to redundant calculations
Solution Approach 1:
The patent divides the prediction calculation process into two hierarchical levels: Level 1 performs basic addition operations (P[n-1]+P[n]+1) that are common to all prediction modes, and Level 2 performs mode-specific operations. This segmentation allows shared computations to be executed once at Level 1, with results reused by multiple prediction modes at Level 2, thereby eliminating redundant calculations while maintaining a relatively simple overall structure.
Solution Approach 2:
The patent merges the computation of multiple prediction modes by having them share common intermediate results from Level 1 operations. Instead of computing each prediction mode independently, the system combines their computational paths by feeding the same Level 1 results to multiple Level 2 processing units, thus reducing total computational operations while serving multiple prediction modes simultaneously.
2Productivity
If calculation bandwidth is increased to compute multiple predicted pixels simultaneously, then productivity improves, but device complexity and chip area increase due to requiring separate datapaths for each pixel
Solution Approach 1:
The patent designs the datapath to be multi-functional, where the same Level 1 and Level 2 processing units can serve multiple prediction modes and multiple pixel positions. By making the computation units universal rather than dedicated to specific modes or pixels, the system achieves high throughput without proportionally increasing hardware resources. The universal datapath can be dynamically configured to handle different prediction modes and pixel positions through control logic rather than dedicated hardware for each case.
Data Source
AI summary
The present invention provides a method of implementing an intra prediction computation applied to the H.264 digital video coding and a device. The method according to present invention of implementing the intra prediction computation applied to the H.264 video coding comprises selecting an image block to be intra-predicted and extracting the neighboring pixel values of said block; determining the prediction mode of said block, and perform first adding operation on said neighboring pixel values when the prediction mode is one of Diagonal_Down_Left, Diagonal_Down_Right, Vertical_Right, Horizontal_Down, Vertical_Left, and Horizontal_Up; performing first shifting operation and second adding operation on the result of first adding operation respectively; performing second shifting operation on the result of second adding operation; selecting a corresponding value to output from the results of first and second shifting operation according to said prediction mode and size of block. The method and circuit device according to present invention will decrease most of the repeated calculations and improve the efficiency without greatly increasing the required chip real estate area.