Elementary Video Bitstream Analysis for Multipoint Conferencing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In multipoint video conferencing, existing systems struggle to effectively analyze and enhance video streams in real-time, leading to suboptimal participant visibility and poor video quality due to factors like camera positioning, lighting, and focus, which are not adequately addressed by current methods.

Innovation Solution

The implementation of elementary bitstream analysis within a multipoint video conferencing unit (MCU) to analyze visual characteristics such as motion, luminance, chrominance, and focus directly from encoded bitstreams without full decoding, allowing for adaptive video composition and feedback to improve video stream quality and participant visibility.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If video streams are fully decoded for analysis, then visual characteristics can be accurately determined, but computational overhead and processing time increase significantly

Engineering Contradiction:
Improvevisual characteristics determination accuracyVSAvoidcomputational overhead
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts only the necessary visual characteristic information directly from the encoded bitstream without performing full decoding. The MCU analyzes motion vectors, macroblock types, and other encoding parameters to determine visual characteristics such as motion activity, scene changes, and region of interest, thereby avoiding the computational burden of complete video decoding while still achieving accurate analysis.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent performs preliminary analysis of visual characteristics on the encoded bitstream before full decoding occurs. By evaluating motion vectors, quantization parameters, and macroblock data in the compressed domain, the system determines which video streams require enhancement or modification in advance, reducing overall processing complexity in the decoding stage.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If all video streams are analyzed and composed, then participant visibility is improved, but processing time and system resources increase

Engineering Contradiction:
Improveparticipant visibilityVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent applies local quality enhancement by identifying specific regions of interest within video streams that require improvement. Rather than uniformly processing all video data, the system analyzes motion vectors and macroblock information to detect regions with poor quality indicators (such as high motion activity, low frame quality, or camera instability) and applies selective enhancement only to those regions, thereby improving participant visibility while minimizing processing time.

Inventive Principle:
Principle #3Local quality

3Manufacturing precision

If video composition is dynamically adjusted based on real-time analysis, then video quality is enhanced, but system complexity increases

Engineering Contradiction:
Improvevideo qualityVSAvoidsystem complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent implements a feedback mechanism where the MCU continuously analyzes encoded bitstreams from multiple video conferencing units, determines visual characteristics and quality metrics, and dynamically adjusts video composition and enhancement parameters based on this feedback. The system monitors motion activity, scene changes, and quality indicators in real-time and automatically modifies stream composition, scaling, and enhancement to optimize video quality while maintaining manageable system complexity through automated control.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3117604B1Elementary video bitstream analysis
Publication Date: 2023.10.04 CISCO TECHNOLOGY INC
  • EP3117604B1 patent drawingFigure 1
  • EP3117604B1 patent drawingFigure 2
  • EP3117604B1 patent drawingFigure 3

AI summary

A bitstream encoded with video information is received at a first network enabled device (105) from a second network enabled device (115a-115c). The bitstream is analyzed prior to generating an image frame corresponding to the bitstream values, and visual characteristics of the video information are determined from bitstream values in the bitstream. A modified video display is determined in response to the visual characteristics.