Video Quality Estimation via TS Packet Header Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video communication techniques struggle to accurately estimate video quality when the payload portion of TS packets is encrypted, as they cannot analyze the encrypted payload to determine frame types, leading to poor video quality due to atmospheric conditions or network issues like packet loss and delay.

Innovation Solution

A video quality estimation apparatus and method that determines frame types based on the number of TS packets in each frame by analyzing the frame start position and data amounts within TS packets, even when the payload is encrypted, using a storage unit and arithmetic processing unit to estimate video quality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If the payload portion of TS packets is encrypted to protect copyright, then security is improved, but frame type analysis becomes impossible leading to poor video quality estimation

Engineering Contradiction:
Improvecopyright protectionVSAvoidframe type analysis accuracy
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The patent extracts frame type information from the TS packet header (specifically the PID and other header fields) rather than from the encrypted payload portion. This allows the system to identify frame types without decrypting the payload, thus maintaining both security and measurement capability.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent uses the TS packet header as an intermediary carrier that contains frame type information. Instead of directly analyzing the encrypted payload, the system uses the unencrypted header fields (PID, program number, etc.) as a mediator to infer frame types, enabling indirect analysis that preserves security.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If frame type information is added to each packet to enable quality analysis, then measurement precision is improved, but device complexity and data processing burden increase

Engineering Contradiction:
Improvequality degradation range graspVSAvoidpacket processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent enables the TS packet structure itself to carry the necessary frame type information through its existing header fields (PID, program number, etc.). The packets essentially provide their own identification information without requiring external addition of frame type markers, reducing processing complexity while maintaining analysis capability.

Inventive Principle:
Principle #25Self-service

3Productivity

If motion prediction encoding is used to compress video, then encoding efficiency is improved, but packet loss affects multiple frames increasing quality degradation range

Engineering Contradiction:
Improveencoding efficiencyVSAvoidvideo quality stability
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent performs preliminary identification of frame types and their spatial-temporal relationships before quality degradation occurs. By pre-mapping which frames depend on which packets (using the frame structure information in the header), the system can predict and compensate for quality degradation effects when packet loss occurs, rather than reacting after degradation has happened.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP2077672B1Video quality estimation device, video quality estimation method, frame type judgment method, and recording medium
Publication Date: 2014.06.25 NIPPON TELEGRAPH & TELEPHONE CORP
  • EP2077672B1 patent drawingFigure 1
  • EP2077672B1 patent drawingFigure 2
  • EP2077672B1 patent drawingFigure 3

AI summary

A frame type determination unit (15B) counts, as the frame data amount of each frame, the number of TS packets included in the frame based on a frame start position included in an input TS packet of video communication, and determines a frame type based on the large/small relationships between the frame data amounts of the frames. A video quality estimation unit (15C) estimates the video quality of the video communication based on the frame type of each frame obtained by the frame type determination unit (15B), the frame structure (14A) of an elementary stream read out from a storage unit (14), and a TS packet loss state detected from the TS packets of the video communication.