Research on Real-time Video Stitching Technology

DOI

10.2991/macmc-17.2018.140How to use a DOI?

Keywords

Image Stitching; Image Registration; Image Composition; Feature Correlation; Panorama

Abstract

Limited by the viewing angles and sizes of imaging instruments like cameras, large pictures can't be generated, and image stitching can solve this problem. image stitching is such a technique that seamlessly mosaics two or more adjacent images with partially overlapping and thus generates a high-definition picture with a relatively wide viewing angle or a panorama with a 360§ viewing angle. This technique involves computer vision, computer graphics, image processing and some mathematical tools and so on. And it makes it possible for noise reduction, field of view (FOV) widening, removal of mobiles, blur removing, spatial resolution improvement and dynamic range enhancement. The paper gives a systemic introduction to the origin, actuality, field of application and mosaic methods and gives a respective and detailed introduction to the two main processes: image registration and image composition. Image registration is the core technique of image stitching. The paper also classifies and summarizes the methods of image registration and described their advantages and disadvantages. At last, it gives a detailed introduction to the common methods of image composition at present and puts forward a feasible method of real-time image stitching.

Copyright

© 2018, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Tao Zou
AU  - Fang Deng
PY  - 2018/01
DA  - 2018/01
TI  - Research on Real-time Video Stitching Technology
BT  - Proceedings of the 2017 4th International Conference on Machinery, Materials and Computer (MACMC 2017)
PB  - Atlantis Press
SP  - 741
EP  - 744
SN  - 2352-5401
UR  - https://doi.org/10.2991/macmc-17.2018.140
DO  - 10.2991/macmc-17.2018.140
ID  - Zou2018/01
ER  -