Multimedia Playback Repositioning via Scene Boundary Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current multimedia playback systems struggle to accurately resume playback from a meaningful point after trickplay operations, often forcing users to apply additional commands to find the desired scene, due to repositioning based on frame boundaries rather than scene boundaries.
Innovation Solution
The system identifies scene boundaries within multimedia content using meta-data and scene change detection, allowing users to resume playback from the start of a scene or the preceding frame, providing a more accurate and user-friendly repositioning mechanism.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If repositioning is based on frame boundaries, then the repositioning mechanism is simple to implement, but the playback may start from anywhere within a scene causing users to miss portions of content
Solution Approach 1:
The patent segments the multimedia content into distinct scenes using scene boundary detection. Instead of treating the entire content as a continuous stream of frames, the system identifies and marks scene boundaries, allowing repositioning operations to align with these semantic segments. This segmentation enables playback to start at meaningful scene boundaries rather than arbitrary frame positions, solving the problem of missing content portions while maintaining implementation feasibility through metadata-based scene identification.
2Device complexity
If repositioning is based on frame boundaries, then the system complexity remains low, but additional trickplay operations are required to find the desired resume point
Solution Approach 1:
The system performs preliminary scene boundary detection and metadata generation during content processing or playback. By pre-identifying scene boundaries and storing this information in metadata, the system prepares the repositioning structure in advance. When a user performs a trickplay operation and subsequently resumes playback, the system can immediately jump to the correct scene boundary without requiring additional user operations to find the desired resume point, thus improving ease of operation while maintaining relatively low system complexity.
3Measurement precision
If playback resumes from the current frame position, then timing accuracy is maintained, but scene accuracy is lost and users may miss important content
Solution Approach 1:
The patent introduces scene boundary metadata as an intermediary layer between the timing information and the actual playback position. Instead of directly resuming from the current frame position (which maintains timing but loses scene accuracy), the system uses scene boundary detection results to determine the optimal resume point. This intermediary metadata structure allows the system to balance both timing accuracy and scene accuracy by selecting scene boundaries as resume points, ensuring users don't miss important content while maintaining reasonable timing precision.
Data Source
AI summary
Methods, systems and computer readable media may facilitate the repositioning of content based upon scene boundaries. In embodiments, content may be automatically repositioned at a scene boundary, so that the content presentation starts from a more meaningful and sensible point after a re-position (e.g., time-shift, trickplay, seek, etc.). Also provided herein is an approach to provide an option for the user to start playback of content from an absolute time based repositioned point or a scene based repositioned point. Scene change detection may be used to detect scene changes within a piece of multimedia content as the content is being received, played, or recorded by a device. Scene changes within a piece of content may be identified using scene meta-data created on a per content basis, wherein the meta-data includes the starting frame offset of each scene along with the duration of the scene.