Video Coding Second Track Sample Metadata Overhead
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing ISO base media file format (ISOBMFF) for video coding and decoding faces inefficiencies in managing Movie Fragment Boxes, leading to significant metadata overhead and non-optimal operation during file reading, especially when using the movie fragment feature for splitting metadata.
Innovation Solution
The method involves generating a second track of samples with multiple access units of the video bitstream, where each sample is encapsulated with metadata such as decoding time, duration, and composition time, and structuring the file according to ISOBMFF for segmented delivery, allowing for optimized delivery of Movie Fragment boxes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the movie fragment feature is used to split metadata into multiple pieces, then the size of the movie box is limited to avoid data loss, but the overhead of the metadata becomes significant compared to the payload
Solution Approach 1:
The patent combines multiple access units into a single sample in the second track, and consolidates metadata associated with these access units into the sample metadata rather than having separate metadata for each access unit. This merging reduces the overall metadata overhead while maintaining the reliability benefits of fragmented metadata storage.
Solution Approach 2:
The sample in the second track serves multiple functions: it contains multiple access units and their associated metadata, acts as a container for segmented delivery, and enables both progressive download and random access. This multi-functionality reduces the need for separate metadata structures for each access unit, thereby reducing overhead.
2Quantity of substance
If movie fragment boxes are compressed by not downloading them altogether but parsing media data, then file size is reduced, but the operation of a file reader becomes non-optimal and unreliable
Solution Approach 1:
The patent pre-organizes multiple access units and their metadata into a structured sample format in the second track before delivery. This preliminary structuring allows file readers to efficiently access metadata without needing to parse through entire media data streams, making the operation both optimal and reliable while maintaining compact file size.
Solution Approach 2:
The second track with its structured samples acts as an intermediary between the compressed movie fragment boxes and the file reader. It provides a standardized interface that enables reliable and efficient metadata access without requiring the file reader to perform complex parsing operations on compressed data.
3Device complexity
If each sample in the first track contains only one access unit, then metadata management is simplified, but the delivery efficiency and segmented delivery capability are reduced
Solution Approach 1:
The patent segments the video content into multiple access units that are grouped into samples in the second track, enabling segmented delivery where portions of the video can be delivered and accessed independently. This segmentation improves delivery efficiency while the structured sample format maintains manageable metadata complexity.
Solution Approach 2:
The patent introduces a second track dimension alongside the first track, where samples in the second track can contain multiple access units. This dimensional addition enables more efficient segmented delivery without increasing metadata management complexity within each individual sample, as the multi-access-unit structure is handled at the track level.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
A method comprising: obtaining video content as an input, wherein said video content comprises an encoded video bitstream or a file comprising a first track of samples, each sample comprising an access unit of the video bitstream; generating a second track of samples associated with the video content, wherein at least one sample of the second track comprises multiple access units of the video bitstream; providing each of the multiple access units within one sample with metadata, wherein said metadata comprises at least one of the following: decoding time, duration, composition time, composition time offset, size, sample flags, sample group description entry of a given sample group; and encapsulating the video content into a file comprising the second track of samples.