Image Mosaic Algorithm of Sequential Images Based on Voronoi

DOI

10.2991/amms-17.2017.10How to use a DOI?

Keywords

component; image mosaicking; moving DLT; bundle adjustment;multi-band blending

Abstract

The image sequence of Vehicle panorama camera platform has the characteristics of serious distortion, large parallax and large amount of data, which can lead to invalid montage for the image sequence by adopting the common image mosaicking method based on features. In this paper, Voronoi algorithm is introduced according to the actual characteristics of vehicle camera platform. A steady and fast image mosaic method for image sequence is proposed. Firstly, the relative positions between images are structured by utilizing GPS information of images. The splicing region of images are classified through Voronoi algorithm and the confidence is introduced to select the best candidate image; then SIFT feature of image is extracted and image registration is carried out by adopting Moving DLT algorithm, what follows is global alignment for stitching images by using improved Bundle Adjustment algorithm. Finally, the stitching images are integrated by using Multi-Band method. Contrasted with experimental results, the proposed method not only realizes the image mosaic under the condition of large scale, multi parallax and distortion images which can make up for the shortcomings of traditional image mosaicking methods, but also reaches good stitching effect and high mosaicking efficiency.

Copyright

© 2017, 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  - Zhicheng Wang
AU  - Qian Xu
AU  - Zhiheng Wang
AU  - Xiaopeng Guo
AU  - Yaxing Yang
PY  - 2017/11
DA  - 2017/11
TI  - Image Mosaic Algorithm of Sequential Images Based on Voronoi
BT  - Proceedings of the 2017 International Conference on Applied Mathematics, Modeling and Simulation (AMMS 2017)
PB  - Atlantis Press
SP  - 45
EP  - 48
SN  - 1951-6851
UR  - https://doi.org/10.2991/amms-17.2017.10
DO  - 10.2991/amms-17.2017.10
ID  - Wang2017/11
ER  -