Live Broadcast Bullet Screen Segmentation for Memory Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current live broadcast playback methods consume excessive memory due to binding all text bullet screens to the recorded live video time axis, leading to poor user immersion and potential playback fluency issues.
Innovation Solution
The method involves segmenting live bullet screen data into multiple segment files based on playback timestamps, allowing for efficient delivery and display of bullet screens during live video playback, reducing memory occupation and improving playback fluency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all text bullet screens are bound to the recorded live video time axis, then users can watch live bullet screens during live broadcast playback, but a large amount of memory is occupied which may cause playback to be stuck and not fluent
Solution Approach 1:
The patent segments the continuous bullet screen data stream into discrete time-axis segments corresponding to different video time ranges. Each segment is independently managed and released when its corresponding video segment is played, preventing accumulation of unnecessary data in memory while ensuring complete bullet screen display during playback.
2Reliability
If all text bullet screens are bound to the recorded live video time axis, then live bullet screens can be watched during playback, but playback fluency deteriorates due to excessive memory occupation
Solution Approach 1:
The patent divides bullet screen data into time-axis segments that are loaded and released in sync with video playback segments. This segmentation enables on-demand loading of only the bullet screens needed for the current playback position, reducing memory pressure and improving playback fluency while maintaining synchronization.
Solution Approach 2:
The patent performs preliminary segmentation of bullet screen data during video encoding, organizing it into time-axis segments before playback. This preliminary organization allows the playback system to efficiently retrieve and display only the necessary bullet screens without excessive memory allocation, ensuring smooth playback.
3Loss of information
If all text bullet screens are bound to the recorded live video, then complete live broadcast scene can be restored, but user sense of immersion deteriorates due to poor playback fluency
Solution Approach 1:
The patent segments bullet screen data into time-axis segments that are efficiently managed during playback. This segmentation ensures complete restoration of the live broadcast scene by preserving all necessary bullet screen information while enabling smooth playback through optimized memory management and on-demand data loading.
Data Source
AI summary
The present application provides techniques of presenting bullet screens for live broadcast playback. The techniques comprise receiving a request for bullet screens sent from a client computing device during playback of at least one portion of a live broadcast video, wherein the bullet screens comprise comments moving across frames of the live broadcast video, the request comprises a timestamp, and the timestamp is associated with a playback progress of the at least one portion of the live broadcast video; determining a bullet screen segment file among a plurality of bullet screen segment files based on the timestamp, wherein the plurality of bullet screen segment files are generated by segmenting live bullet screen data associated with the live broadcast video into N bullet screen segment files, and the N is a positive integer; and transmitting the bullet screen segment file to the client computing device for display of the bullet screens.


