Video Frame Encoding with Non-Adjacent Redundancy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In videoconferencing systems, packet loss during network congestion leads to the potential loss of both primary and redundant frames, limiting transmission flexibility and robustness due to the requirement for adjacent decoding of primary and redundant frames.
Innovation Solution
The method involves encoding and decoding video data by including indication and location information in bitstreams to allow a gap between primary and redundant frames, enabling flexible transmission and reducing the likelihood of frame loss, while maintaining robustness by carrying location relationship information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If primary frame and redundant frame are transmitted adjacently in the bitstream, then decoding simplicity is improved, but transmission robustness deteriorates due to increased risk of losing both frames simultaneously
Solution Approach 1:
The patent segments the bitstream by inserting a network abstraction layer unit between the primary frame and redundant frame. This segmentation separates two previously adjacent frames into distinct transmission units, allowing them to be independently handled during transmission and reducing the risk that network operations affect both frames simultaneously.
Solution Approach 2:
The patent introduces a network abstraction layer unit as an intermediary element between the primary frame and redundant frame. This intermediary carries identification information and serves as a buffer or separator, preventing direct adjacency while enabling the decoder to properly identify and process the frames according to their types.
2Ease of operation
If primary frame must arrive before redundant frame, then decoding order is simplified, but transmission flexibility deteriorates
Solution Approach 1:
The patent makes the decoding process dynamic by enabling the decoder to adapt its behavior based on the actual reception status of frames. The network abstraction layer unit carries identification information that allows the decoder to dynamically adjust whether to use the primary frame, redundant frame, or both, depending on which frames were successfully received, rather than following a fixed decoding order.
Solution Approach 2:
The patent implements a feedback mechanism where the decoder uses the identification information in the network abstraction layer unit to determine the appropriate decoding strategy. The decoder receives feedback about frame types and arrival status, then adjusts its decoding behavior accordingly - using the primary frame if available, falling back to the redundant frame if needed, or combining both for enhanced robustness.
3Device complexity
If no location information is carried for frames, then data structure is simpler, but ability to handle non-adjacent frames deteriorates
Solution Approach 1:
The patent applies preliminary action by pre-encoding location information and frame type identification into the network abstraction layer unit before transmission. This allows the decoder to know in advance the types and locations of frames without requiring complex runtime analysis, enabling efficient handling of non-adjacent frame structures while maintaining relatively simple decoding logic.
Data Source
AI summary
Video data encoding and decoding methods and apparatuses which include determining a to-be-encoded information group in a to-be-encoded video sequence, generating a primary information group corresponding to the to-be-encoded information group, wherein the primary information group comprises a first data information, indication information that indicates whether the to-be-encoded information group is the primary information group or a redundant information group, and location information of the primary information group, generating the redundant information group corresponding to the to-be-encoded information group, wherein the redundant information group comprises a second data information, location information of the redundant information group, indication information that indicates whether the to-be-encoded information group is the primary information group or the redundant information group, and location relationship information of the redundant information group and the primary information group, and encoding the primary information group and the redundant information group into a bitstream.