Live Video Streaming Bandwidth Reduction via Object Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for reducing network bandwidth during live video sharing, such as resolution reduction techniques, often obscure important details in video frames, making them unsuitable for public safety applications like car chases or suspect pursuits, where clear visualization of license plates or facial features is crucial.

Innovation Solution

The system processes video streams by initially transmitting full frames with metadata indicating objects of interest and movement profiles, allowing the receiving device to extract and transform background portions, thereby reducing bandwidth consumption while maintaining high-quality video by streaming only object of interest portions and rendering transformed backgrounds in real-time.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of energy

If resolution reduction techniques are applied to reduce network bandwidth, then network bandwidth consumption is reduced, but important details in video frames are obscured

Engineering Contradiction:
Improvenetwork bandwidth consumptionVSAvoidimportant video details
Core Design Contradiction:
Loss of energyVSLoss of information

Solution Approach 1:

The video frame is segmented into two distinct portions: an object of interest portion containing critical details (license plates, facial features) and a background portion. This segmentation allows the system to transmit only the essential object information at high resolution while reducing overall bandwidth consumption by excluding redundant background data.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system extracts and transmits only the object of interest portion from each video frame, separating it from the background. This extraction technique removes unnecessary background information from the transmission stream, reducing bandwidth requirements while preserving all critical details about objects of interest.

Inventive Principle:
Principle #2Taking out (Extraction)

2Loss of energy

If uniform resolution reduction is applied across all video frames, then network bandwidth is reduced, but critical details such as license plates and facial features become obscured

Engineering Contradiction:
Improvenetwork bandwidth consumptionVSAvoiddetail clarity
Core Design Contradiction:
Loss of energyVSMeasurement precision

Solution Approach 1:

Different quality levels are applied to different parts of the video frame. The object of interest portion maintains high resolution and quality to ensure clear visualization of critical details, while the background portion is excluded from transmission or reduced to minimal data. This local quality differentiation resolves the contradiction between bandwidth reduction and detail preservation.

Inventive Principle:
Principle #3Local quality

3Loss of energy

If only object of interest portions are transmitted, then network bandwidth consumption is reduced, but the receiving device must perform additional processing to reconstruct complete frames

Engineering Contradiction:
Improvenetwork bandwidth consumptionVSAvoidvideo processing complexity
Core Design Contradiction:
Loss of energyVSDevice complexity

Solution Approach 1:

The electronic image capturing device performs preliminary actions by extracting and separating the object of interest portion from the background portion before transmission. This preliminary segmentation reduces the amount of data that needs to be processed at the receiving device, as only the essential object information is transmitted rather than complete frames requiring full reconstruction.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3867859B1Systems and methods for processing a video stream during live video sharing
Publication Date: 2024.11.27 MOTOROLA SOLUTIONS INC
  • EP3867859B1 patent drawingFigure 1
  • EP3867859B1 patent drawingFigure 2A
  • EP3867859B1 patent drawingFigure 2B

AI summary

Systems and methods for processing video stream data during live video sharing may reduce network bandwidth consumption by streaming full and partial video frames. An image processing device receives information from an image capturing device including a first full video frame, an object of interest identifier, and a movement profile of the image capturing device. The image processing device extracts a background portion from the first full video frame, exclusive of the object of interest. Subsequently, the image processing device receives a partial second video frame including only an object of interest portion, applies a transformation to the extracted background portion based on the movement profile, and displays a representation of the second video frame including the object of interest portion of the second video frame and the transformed background portion. The image processing device performs fine tuning to compensate for small background or movement profile changes between partial frames.