Video Encoding Slice Division for Parallel Processing Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional moving picture encoding/decoding apparatuses that use parallel processing by dividing images into slices suffer from degraded encoding performance compared to encoding the entire image, necessitating a more efficient slice division method to enhance encoding efficiency.
Innovation Solution
A moving picture encoding/decoding apparatus and method that divides input images or bitstreams into slices, with a slice divider and synchronization controller to determine encoding/decoding orders, allowing for parallel processing and information sharing between slices to improve encoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If parallel processing is implemented by dividing images into slices, then processing speed and productivity are improved, but encoding performance and quality deteriorate
Solution Approach 1:
The image is divided into multiple slices that can be processed in parallel by different threads or cores. Each slice is independently encoded while maintaining the ability to share information with adjacent slices through boundary macroblocks, thus achieving both parallel processing efficiency and encoding quality.
Solution Approach 2:
Boundary macroblocks between adjacent slices serve as intermediaries that enable information sharing between parallel processing units. These boundary regions allow motion estimation and compensation data to be exchanged between slices, maintaining encoding accuracy while enabling parallel processing.
2Manufacturing precision
If motion estimation is performed with forward/backward reference frames, then encoding quality is improved, but computational complexity increases exponentially
Solution Approach 1:
The computational task of motion estimation is segmented across multiple parallel processing units. Each thread or core handles motion estimation for a specific slice independently, reducing the computational burden on single processors while maintaining overall encoding quality through reference frame utilization.
Solution Approach 2:
Motion estimation results and reference frame data are prepared and stored in advance in memory buffers before the actual encoding process. This preliminary preparation allows parallel processing units to access pre-computed motion data without increasing real-time computational complexity during the encoding phase.
Data Source
AI summary
Disclosed is a moving picture encoding/decoding apparatus and method for processing of a moving picture, which is divided in units of slices. The encoding method includes the steps of: dividing a moving picture in units of slices when the moving picture is received; determining an encoding order of moving pictures divided in units of slices; and generating a bitstream by encoding moving pictures, which are divided in unites of slices, according to a corresponding order when the encoding order has been determined.