Dynamic Video Stitching Without Transcoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Content providers face challenges in dynamically inserting videos, such as advertisements, into hosted videos without control over the video player, particularly due to the high computational costs associated with transcoding large sets of videos and the staleness of stitched videos.
Innovation Solution
A system and method for dynamically stitching an insert video into a target video without transcoding, allowing for real-time updates of the insert video content, such as advertisements, by determining insertion points and updating sample tables, while transmitting the stitched video over HTTP.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If transcoding is used to stitch video content, then video stitching can be achieved, but computational costs become expensive
Solution Approach 1:
The patent segments the video processing task by separating the host video and insert video into independent encoded streams. Instead of transcoding the entire stitched video, the system processes only the metadata and timing information to combine the pre-encoded video segments, dramatically reducing computational requirements while maintaining stitching functionality.
2Adaptability or versatility
If transcoding is used to stitch videos, then video content can be dynamically inserted, but the stitched video becomes stale quickly
Solution Approach 1:
The patent implements dynamic video stitching by allowing the insert video to be changed without reprocessing the host video. The system uses metadata-driven assembly where the host video remains static and pre-encoded, while insert videos can be dynamically swapped by updating only the stitching metadata, enabling real-time content updates without full transcoding.
3Ease of operation
If a content provider does not control the video player, then streaming flexibility is reduced, but transcoding requirements increase
Solution Approach 1:
The patent introduces an intermediary metadata layer that acts as a bridge between the content provider and the uncontrolled video player. The metadata contains timing, synchronization, and assembly instructions that enable the client-side player to perform the stitching operation locally without requiring server-side transcoding or player control, thus maintaining compatibility with standard players while enabling dynamic content insertion.
Data Source
AI summary
Systems and methods are provided herein relating to video editing and more particularly to stitching an insert video within a target video without transcoding. Through dynamically stitching a video, such as an advertisement, within a video, a content provider can transmit a stitched video instead of separate content videos and advertisement videos that a local uncontrolled video player would be responsible for combining and playing. Systems and methods herein provide for receiving a target video and an insert video and dynamically stitching the insert video within the target video to create a stitched video. The stitched video can then be transmitted that plays both the target video and the insert video within the target video, irrespective of the player on which a user views the stitched video.