Parallel Image Encoding via Motion Vector Timing Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image encoding systems face inefficiencies when performing parallel encoding of multiple images due to dependency on specific encoded streams, leading to reduced processing performance and increased data size, as they often rely on temporally adjacent images for inter-picture encoding, which deteriorates compression efficiency.
Innovation Solution
An image encoding apparatus that encodes images block-by-block, using a control unit to manage the start timing of encoding units based on the presence of motion vectors, allowing for parallel processing without relying on specific encoded streams, by instructing subsequent units to start encoding only after a certain number of block lines including motion vector ranges are processed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a plurality of image encoding apparatuses is used to perform parallel encoding processing, then processing performance is improved, but dependency on specific encoded streams causes encoding to stall when reference images are being processed by other apparatuses
Solution Approach 1:
The patent introduces an encoded stream buffer as an intermediary component between the parallel encoding apparatuses. This buffer stores encoded streams and makes them available as reference images without requiring direct dependency on other encoding apparatuses being complete, thereby enabling parallel processing while maintaining reference image availability.
Solution Approach 2:
The patent segments the encoding process into independent parallel apparatuses, each capable of encoding different images simultaneously. By using the buffer to decouple the dependency relationships, each apparatus can operate independently on its assigned images without waiting for others to complete their reference image processing.
2Productivity
If a temporally distant image is used as a reference image for parallel encoding, then encoding dependency is reduced, but compression efficiency deteriorates and data size increases
Solution Approach 1:
The patent performs preliminary encoding and stores the encoded streams in a buffer before they are needed as reference images. This allows the encoding process to proceed in parallel without waiting for reference images to be ready, while still using temporally adjacent images for optimal compression when they become available.
Data Source
AI summary
An image encoding apparatus includes a first encoding unit configured to encode an N-th image of a plurality of images and a second encoding unit configured to encode an (N+1)th image, which is continuous to the N-th image, with reference to the N-th image. The processing system adaptively determines processing start timing based on a motion vector presence range of an initial processing target block of the (N+1)th image, which is a part of the N-th image.