Automated Detection of 3D Video Formats for Mixed Reality Headsets: A Feature Matching Approach Using Epipolar Geometry and Disparity Analysis
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
In the context of mixed reality technologies, the need for accurately detecting 3D video projection formats, such as side-by-side and top-bottom, is critical for providing seamless and immersive experiences. This paper presents an automated algorithm for identifying 3D video formats using feature-matching techniques, including ORB for keypoint detection and RANSAC for fundamental matrix estimation, combined with disparity analysis. Integrated into a video player, the tool processes video frames in real-time, classifying them as 3D SBS, 3D TB, or non-3D. The proposed method achieved an accuracy of 97.89% when tested on 3,474 frames across diverse video formats. The algorithm offers an efficient solution for enhancing user experiences in MR environments, particularly for users unfamiliar with 3D content. Future work may focus on expanding its applicability to live video streams and further improving classification precision.
Fields of Study & Subjects
Advanced Vision and Imaging
Advanced Optical Imaging Technologies
Video Coding and Compression Technologies
Publication details
- Journal / source
- International Conference on Consumer Electronics
- Pages
- 1–6
- Publication date
- 3 February 2026
- Language
- EN