Video Substream Playback via SEI Slice Extraction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional video playing solutions face server overload issues due to high I/O and bandwidth pressures, leading to lag problems when multiple clients request video data simultaneously, and attempting to alleviate this by downloading video data from alternative sources increases delay times.

Innovation Solution

A video processing method that acquires and plays video substreams through multiple channels, extracts video slice numbers from Supplemental Enhancement Information (SEI) frames, determines missing slices, and downloads them from corresponding sources to maintain seamless playback, thereby reducing server load and delay.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If video data is buffered in conventional solutions to avoid lag, then playback smoothness is improved, but server I/O pressure and bandwidth consumption increase

Engineering Contradiction:
Improveplayback smoothnessVSAvoidserver I/O pressure
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The video stream is segmented into multiple substreams (first video substream from server, second video substream from peer terminal) that can be downloaded simultaneously through different channels. This segmentation allows the system to distribute the download load across multiple sources, reducing pressure on any single server while maintaining continuous playback.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

A peer terminal acts as an intermediary video source. Instead of all clients directly downloading from the origin server, video data can be obtained from peer terminals that have already downloaded the data. This intermediary approach reduces origin server I/O pressure while maintaining data availability for continuous playback.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Quantity of substance

If video data is downloaded from alternative video sources to relieve server pressure, then server I/O pressure is reduced, but video playing delay increases

Engineering Contradiction:
Improveserver I/O pressureVSAvoidvideo playing delay
Core Design Contradiction:
Quantity of substanceVSLoss of time

Solution Approach 1:

The system performs preliminary actions by extracting and analyzing SEI frame information to identify missing video slices before they cause playback interruption. By proactively detecting missing slices through SEI frame analysis and immediately downloading them from alternative sources, the system avoids playback delays while reducing server pressure.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses SEI frames as feedback mechanisms to monitor video stream completeness. By continuously analyzing SEI frame information to detect missing video slices and immediately responding by downloading the missing data, the system creates a feedback loop that maintains playback continuity while utilizing alternative sources to reduce server load.

Inventive Principle:
Principle #23Feedback

3Productivity

If multiple video substreams are acquired through multiple channels, then server pressure is reduced and delay is minimized, but system complexity increases

Engineering Contradiction:
Improvevideo download efficiencyVSAvoidsystem complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The system extracts only the necessary information (video slice numbers) from SEI frames rather than processing entire video streams. This extraction approach simplifies the complexity of handling multiple video substreams by focusing only on the critical metadata needed to identify and download missing slices, while still achieving the benefits of multi-channel acquisition.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentUS12058387B2Video processing method and apparatus, computer device, and storage medium
Publication Date: 2024.08.06 TENCENT TECHNOLOGY (SHENZHEN) CO LTD
  • US12058387B2 patent drawing
  • US12058387B2 patent drawing
  • US12058387B2 patent drawing

AI summary

A video processing method includes: acquiring video substreams through at least two channels, to obtain at least two video substreams each having a one-to-one correspondence with one of the at least two channels (S202); extracting video slice numbers from supplemental enhancement information (SEI) frames of the at least two video substreams (S204); determining a missing video slice missed from a target video substream of the at least two video substreams according to the extracted video slice numbers (S206); downloading the missing video slice from a video source corresponding to the target video substream (S208); and playing the downloaded video slice and the at least two video substreams according to a sequence of the video slice numbers (S210).