Video Frame Type Detection Without Payload Decoding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for detecting frame types in video transmission are inaccurate, particularly in real-time applications, due to reliance on reduction factors and human intervention, and fail to consider global features, leading to incorrect identification of I-frames and high error rates.

Innovation Solution

A method that determines frame types based on the coding sequence and comparison of data amounts without decoding the payload, using dynamic thresholds and automatic parameter updates to improve accuracy and robustness, allowing for real-time detection of I-frames, P-frames, and B-frames.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If frame type detection uses reduction factors and manual intervention, then detection can be performed, but detection accuracy is low and error rate is high

Engineering Contradiction:
Improveframe type detection accuracyVSAvoiddetection reliability
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The system automatically updates parameters and thresholds based on incoming video data without manual intervention. The detector self-adjusts by analyzing statistical features of the video stream and dynamically modifying detection parameters, eliminating the need for manual parameter tuning and reducing human intervention while improving accuracy and reliability

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The detection system transitions from static parameters to dynamic parameter adjustment. Thresholds and parameters are continuously updated based on real-time video stream analysis, allowing the system to adapt to varying video content and conditions, thereby improving both accuracy and reliability across different video scenarios

Inventive Principle:
Principle #15Dynamics

2Productivity

If frame type detection considers only local features, then detection speed is fast, but global features are not considered leading to incorrect I-frame identification

Engineering Contradiction:
Improvedetection speedVSAvoidI-frame identification accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The system merges local feature analysis with global feature consideration. It combines statistical features from the video stream with detection results, integrating multiple levels of information (local packet characteristics and global video structure) to improve I-frame identification accuracy while maintaining detection speed through efficient feature fusion

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The system adds a global dimension to the detection process by considering statistical features and parameter updates across the entire video stream, not just local packet characteristics. This multi-dimensional approach (combining local packet analysis with global stream statistics) enables accurate I-frame identification without sacrificing detection speed

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Measurement precision

If frame type detection requires decoding payload, then detection can be performed accurately, but real-time processing is not achieved

Engineering Contradiction:
Improveframe type detection accuracyVSAvoidreal-time processing capability
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The system extracts and uses only the necessary statistical features and metadata from the video stream without requiring full payload decoding. By extracting key characteristics (packet lengths, timestamps, sequence numbers) and performing detection based on these extracted features alone, the system achieves both accurate frame type detection and real-time processing capability

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system performs partial decoding by analyzing only the essential statistical features and metadata needed for frame type detection, rather than fully decoding the entire payload. This partial action approach provides sufficient information for accurate detection while maintaining real-time processing speed

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP2637410B1Detection method and device for frame type
Publication Date: 2020.09.09 HUAWEI TECH CO LTD
  • EP2637410B1 patent drawingFigure 1A~1B
  • EP2637410B1 patent drawingFigure 2a~2b
  • EP2637410B1 patent drawingFigure 3

AI summary

Embodiments of the present invention disclose a method and an apparatus for detecting frame types. The method includes: detecting play time of each frame; and if the play time of a current frame is less than the maximum play time of received frames, determining the current frame as a bi-directional predicted frame, B-frame. In the technical solutions provided by the embodiments of the present invention, the frame types are determined according to a coding sequence of various frames and comparison between data amounts of various frames received earlier and later, without decoding a payload. Therefore the impact of a reduction factor is eliminated and the frame types can be detected more accurately.