Seekable Video Recording via Pre-calculated Keyframes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Compressed video datasets encoded with Advanced Video Coding (AVC) like H.264 require time-consuming rendering processes when seeking specific frames, and may experience desynchronization between video and audio datasets due to the presence of few fully-coded frames and many partially-coded frames.
Innovation Solution
A recording server that replaces partially-coded frames with fully-coded frames in the video dataset, while maintaining synchronization with audio data, to create a seekable media file by calculating corresponding fully-coded frames and allocating serial numbers for each frame based on timestamps, ensuring frames are correctly placed in time slots.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If video streams are compressed using AVC/H.264 with reference frame encoding, then storage and bandwidth volume are reduced, but seeking specific frames becomes time-consuming and resource-intensive
Solution Approach 1:
The patent pre-calculates and stores fully-coded replacement frames during the recording phase, so that when seeking is needed, the pre-computed frames are immediately available without requiring time-consuming rendering operations. This transforms a post-processing computational burden into a pre-processing storage investment.
Solution Approach 2:
The patent creates copies of fully-coded frames by adding pixel changes from partially-coded frames to the reference frame. These copied fully-coded frames are stored in the video dataset, enabling direct seeking without rendering. The copy operation preserves the visual information while making it immediately accessible.
2Quantity of substance
If most frames are partially-coded to reduce storage volume, then bandwidth and storage efficiency improve, but video and audio desynchronization occurs during playback
Solution Approach 1:
The patent selectively replaces only certain partially-coded frames with fully-coded frames, rather than converting all frames. This local quality change ensures that keyframes and important transition points are fully-coded for synchronization stability, while other frames remain compressed for efficiency.
Solution Approach 2:
The patent performs frame replacement during the recording and encoding phase, establishing proper synchronization before playback. By pre-processing the video dataset to ensure adequate fully-coded frames are present, the system avoids synchronization issues during actual playback without requiring real-time processing.
3Speed
If all frames are fully-coded to enable fast seeking, then seeking speed improves, but storage and bandwidth requirements increase significantly
Solution Approach 1:
The patent applies different coding qualities to different frames based on their importance and position in the sequence. Critical frames that serve as seeking targets are fully-coded, while intermediate frames remain partially-coded. This selective approach enables fast seeking at important points without the storage overhead of fully-coding every frame.
Solution Approach 2:
The patent uses a partial approach by replacing only a subset of partially-coded frames with fully-coded frames, rather than converting all frames. This partial action provides sufficient seeking capability for practical purposes while maintaining storage efficiency, avoiding the excessive storage requirements of complete full-coding.
Data Source
AI summary
A recording server and method for recording and modifying a video dataset, the recording server including a memory storing a code, and at least one processor configured to execute the code, wherein the code includes instructions for receiving a video dataset including an initial fully-coded frame and subsequent plurality of partially-coded frames, selecting at least one of the plurality of partially-coded frames, for replacement of each of the selected frames with a corresponding fully-coded frame, and for each selected partially-coded frame, calculating the corresponding fully-coded frame and replacing the selected partially-coded frame with the calculated corresponding fully-coded frame.