Parallel Video Encoding for Real-Time 360-Degree VR Stitching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 360-degree VR video technology is inefficient in converting large-sized videos into panoramic videos in real time, failing to provide a seamless and immersive VR experience.
Innovation Solution
A video conversion apparatus and system that processes multiple videos in real time by using a controller with a frame capture unit, frame encoder unit, and video stitching unit to encode and stitch video information in a parallel structure, utilizing capture cards and FFMPEG for encoding and calibrating frames from multiple cameras to generate a 360-degree VR video.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional video conversion methods are used to convert large-sized videos into 360-degree VR videos, then the conversion quality can be maintained, but the conversion time becomes excessively long and real-time processing is not achieved
Solution Approach 1:
The patent divides the video conversion process into multiple independent parallel encoding streams, where each camera's video is encoded separately simultaneously. This segmentation allows the system to process multiple video sources concurrently rather than sequentially, dramatically improving conversion speed while maintaining quality
Solution Approach 2:
The patent transitions from traditional sequential processing to a parallel processing architecture by adding the dimension of concurrent execution. Multiple encoders work simultaneously on different video streams, and the stitching unit combines these parallel outputs, achieving real-time conversion capability
2Adaptability or versatility
If multiple cameras are used to capture 360-degree videos, then the panoramic coverage is improved, but the complexity of encoding and stitching multiple video streams increases
Solution Approach 1:
The patent employs a universal stitching unit that can handle multiple video streams from different cameras using the same processing logic and algorithms. This multi-functional stitching unit processes all camera inputs through a unified framework, managing complexity while supporting panoramic coverage
Solution Approach 2:
The patent merges multiple encoded video streams into a single unified 360-degree output through the stitching unit. By combining the parallel encoding results and synchronizing them temporally and spatially, the system achieves comprehensive panoramic coverage while managing complexity through integrated processing
Data Source
AI summary
Disclosed are a video conversion apparatus and a system for generating a 360-degree VR video in real time. The video conversion apparatus includes an input unit and a controller. The input unit receives a plurality of pieces of video information having continuity from a video recording apparatus including a plurality of cameras. The controller captures the plurality of pieces of video information received from the input unit, encodes the captured pieces of video information in a parallel structure, and stitches the encoded pieces of video information to generate a 360-degree VR video.