Block-Based Video Decoding Using Prescan Information
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video decoding methods face inefficiencies in processing pipeline decoding of images, particularly when dealing with high-resolution and high-quality images, leading to increased latency and decoding time due to the need for time synchronization and prediction across picture units, especially in environments with advanced hardware capabilities.
Innovation Solution
A block-based video decoding method and apparatus that utilizes prescan information to enable raster scan decoding processes according to intra-picture block order, independent of coding units, facilitating decoding and pipeline processing by generating prescan information from bitstreams and using it for sequential block decoding within the decoding module.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If picture units are decoded independently by tile division in raster scan order, then decoding can be performed in a structured manner according to coding units, but processing pipeline efficiency is reduced and latency increases due to time synchronization requirements
Solution Approach 1:
The picture is divided into multiple tiles, and each tile is further divided into coding units (CUs) of various sizes using a quad-tree structure. This segmentation allows independent processing of each tile while maintaining overall picture coherence, enabling parallel pipeline processing without requiring time synchronization across the entire picture.
Solution Approach 2:
The invention introduces a prescan process that performs preliminary analysis of the bitstream to generate prescan information about coding unit positions and sizes before actual decoding. This preliminary action enables the decoding pipeline to be properly configured in advance, allowing seamless picture-wide raster scan decoding without time synchronization delays.
2Stability of the object's composition
If sequential decoding is performed for each tile in raster scan order, then decoding structure is maintained, but decoding time increases and latency is increased in pipeline decoding
Solution Approach 1:
The prescan process performs preliminary analysis to generate prescan information about all coding units in the picture before actual decoding begins. This includes identifying CU positions, sizes, and boundaries across tile boundaries. By having this information prepared in advance, the decoding pipeline can proceed seamlessly without time synchronization delays, reducing overall decoding time and latency while maintaining structured decoding.
Solution Approach 2:
The prescan information acts as an intermediary data structure that bridges the tile-based coding structure and the picture-wide raster scan decoding requirement. It provides the necessary metadata about CU boundaries and positions that allows the decoder to maintain structural consistency while achieving seamless picture-wide processing without time synchronization overhead.
3Productivity
If picture-based pipeline decoding is implemented with tile division, then hardware performance can be improved through buffering, but time synchronization and prediction across tiles become difficult and increase processing time
Solution Approach 1:
The picture is segmented into independent tiles that can be buffered and processed in parallel by hardware. Each tile contains complete coding unit information that can be independently decoded while maintaining picture-wide raster scan order through the prescan data structure. This segmentation enables hardware buffering and parallel processing without requiring complex time synchronization mechanisms.
Solution Approach 2:
The prescan information serves as an intermediary that eliminates the need for complex time synchronization and prediction mechanisms. By providing pre-computed CU position and boundary data, it allows the hardware pipeline to process tiles independently while maintaining correct decoding order, significantly reducing device complexity compared to approaches requiring real-time synchronization across tile boundaries.
Data Source
AI summary
Provided are a block based video decoding method and an apparatus thereof, the method includes receiving a bitstream; generating a prescan information corresponding to a picture unit image from the bitstream; and decoding the picture unit image of the bitstream using the prescan information. The bitstream includes the image information in which the picture unit image is divided into one or more coding unit and sequentially block-coded according to a predetermined order in the coding unit, and the decoding comprising processing a block decoding in accordance with the predetermined order in the intra-picture in a non-dependent manner with an encoding order in an intra-coding unit using the prescan information.