Adaptive VR Streaming via DASH Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The high bitrate and resolution of stitched UHD video streams for virtual reality media data streams pose challenges in bandwidth efficiency and processing capabilities, necessitating optimized access and delivery methods.

Innovation Solution

The implementation of adaptive streaming using the DASH protocol with new VR descriptors in the media presentation description (MPD) to signal characteristics like projection type, viewpoint, and region accessibility, allowing clients to select appropriate content and interface settings for efficient data retrieval and rendering.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If stitched UHD video streams are used for virtual reality media data streams, then immersive experience and video quality are improved, but bandwidth consumption and processing requirements increase significantly

Engineering Contradiction:
Improvevideo qualityVSAvoidbandwidth consumption
Core Design Contradiction:
Manufacturing precisionVSQuantity of substance

Solution Approach 1:

The panoramic video stream is divided into multiple independent rectangular video segments corresponding to different regions of interest. Each segment can be independently encoded, transmitted, and decoded, allowing selective delivery of only the segments needed for the current viewpoint, thereby reducing overall bandwidth consumption while maintaining UHD quality for displayed regions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different regions of the panoramic video are encoded with different quality levels and resolutions based on their importance and likelihood of being viewed. Regions that are more likely to be in the user's field of view are encoded with higher quality, while less important regions use lower quality, optimizing the trade-off between video quality and bandwidth usage.

Inventive Principle:
Principle #3Local quality

2Manufacturing precision

If stitched UHD video streams are used for virtual reality media data streams, then immersive experience and video quality are improved, but processing capabilities required increase significantly

Engineering Contradiction:
Improvevideo qualityVSAvoidprocessing capabilities
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

By segmenting the panoramic video into smaller rectangular regions, the decoding and rendering process is simplified. The client device only needs to decode and render the specific segments corresponding to the current viewpoint rather than processing the entire high-resolution panoramic stream, significantly reducing processing requirements while maintaining UHD quality for the displayed portions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system encodes the complete panoramic video at UHD quality, but the client device only processes and displays a partial subset of segments corresponding to the current field of view. This allows the server to prepare high-quality content without requiring the client device to have the processing power to handle the entire panoramic stream at full resolution.

Inventive Principle:
Principle #16Partial or excessive action

3Device complexity

If traditional streaming methods are used without adaptive protocols, then implementation simplicity is maintained, but adaptability to different network conditions and device capabilities is reduced

Engineering Contradiction:
Improveimplementation simplicityVSAvoidadaptability to network conditions
Core Design Contradiction:
Device complexityVSAdaptability or versatility

Solution Approach 1:

The streaming system dynamically adapts the delivered video segments based on real-time network conditions and client device capabilities. The server can adjust segment resolution, quality, and selection based on available bandwidth, while the client can request specific segments based on current viewpoint and device processing power, creating a flexible adaptive streaming solution.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The segmented panoramic video approach works across multiple device types and network conditions. The same segmented structure can be delivered to high-end devices with full UHD quality or adapted for mobile devices with lower resolutions, making the system universally applicable while maintaining adaptability to different scenarios.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP3466093B1Method, device, and computer program for adaptive streaming of virtual reality media content
Publication Date: 2021.12.01 CANON KK
  • EP3466093B1 patent drawingFigure 1a~1e
  • EP3466093B1 patent drawingFigure 2a~2c
  • EP3466093B1 patent drawingFigure 3~5

AI summary

The invention relates to the streaming of media data representing a capturing projection of a wide view of a scene, the streamed media data making it possible to render the wide view on a 3D geometric display surface or to render the wide view on a display surface according to different viewpoints, the rendering comprising rendering projection of media data. After having received a description file including information on the media data, the information comprising descriptive information relating to the capture of the wide view for producing the media data, request messages to request media data streams based on the description file are sent to the server. In response to the request messages, media data corresponding to the requested media data streams are received.