Video Coding Using Composite Depth and Texture Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding methods face reduced efficiency when coding composite pictures formed by stitching texture and depth images from multiple viewports, due to differences in characteristics between texture and depth images, which limits the application of encoders and decoders with specific sampling formats.
Innovation Solution
A method and system for video coding that generates and decodes video data using composite depth and texture images, allowing for improved encoding efficiency by converting depth images to a compatible format and arranging texture and depth images from multiple cameras, enabling efficient compression and decoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If existing video encoding methods are used to code composite pictures formed by stitching texture and depth images from multiple viewports, then encoding can be performed, but encoding efficiency is reduced due to differences in characteristics between texture and depth images
Solution Approach 1:
The patent separates the encoding process into two distinct stages: first encoding depth images using a depth-optimized encoder with YUV400 sampling format, then encoding texture images using a texture-optimized encoder with YUV420 sampling format. This segmentation allows each encoder to be specialized for its specific image type, resolving the contradiction between encoding efficiency and format compatibility by eliminating the need for a single encoder to handle both image types with different characteristics.
Solution Approach 2:
The patent changes the sampling format parameter differently for depth and texture images: depth images use YUV400 format while texture images use YUV420 format. This parameter change optimizes encoding efficiency for each image type according to its specific characteristics, while maintaining overall system compatibility through the unified composite picture structure.
2Productivity
If depth images are encoded using YUV400 sampling format and texture images using YUV420 sampling format, then encoding efficiency is improved, but device complexity increases due to requiring multiple encoders and decoders
Solution Approach 1:
The patent merges the separately encoded depth and texture images into a unified composite picture structure during the encoding process. By combining YUV400 depth images and YUV420 texture images into a single composite picture with a defined data structure, the system maintains encoding efficiency benefits while presenting a unified interface that reduces the practical impact of having multiple specialized encoders.
Solution Approach 2:
The composite picture structure serves multiple functions: it containers both depth and texture images with different sampling formats, provides a unified interface for transmission and storage, and enables flexible extraction and processing of individual image types. This multi-functionality reduces device complexity by allowing a single composite picture handling pipeline to manage multiple image types.
3Adaptability or versatility
If composite pictures are formed by stitching texture and depth images from multiple viewports, then immersive video functionality is enabled, but encoding efficiency is reduced due to characteristic differences between image types
Solution Approach 1:
The patent performs preliminary encoding of depth and texture images separately into their optimized formats (YUV400 and YUV420 respectively) before combining them into the composite picture structure. This preliminary action ensures that each image type is already optimized for its specific characteristics before being stitched together, preventing encoding efficiency loss that would occur if they were combined first and then encoded together.
Solution Approach 2:
The patent creates a composite picture structure that functions like a composite material, combining depth image data and texture image data in a unified format while preserving their individual optimized encoding characteristics. This composite structure enables immersive video functionality by integrating multiple viewport images with different properties, while maintaining encoding efficiency through the use of format-appropriate encoding for each component.
Data Source
AI summary
A method, computer program, and computer system is provided for video coding. Video data including one or more views is received. A composite depth image and a composite texture image corresponding to the one or more views are generated based on the received video data. The video data is decoded based on the generated composite depth image and composite texture image.