Open AccessParallax-Tolerant Image Stitching for Outdoor Scenes
Author(s) -
Chunbin Li,
Jiahang Liu
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1651/1/012186
Subject(s) - image stitching , panorama , computer vision , artificial intelligence , computer science , homography , parallax , image registration , similarity (geometry) , feature (linguistics) , computer graphics (images) , transformation (genetics) , image (mathematics) , mathematics , linguistics , statistics , philosophy , biochemistry , projective test , chemistry , projective space , gene
Image stitching aims to obtain a natural panorama without ghosts. This paper proposes a panoramic image stitching method for outdoor scenes. This method gets the new homography matrix by analyzing the spatial distribution of the feature points of the outdoor scene image, and then combines with the local registration to achieve precise registration. In order to generate a more natural panorama, this paper also combines the result of registration with local similarity transformation and seam-cutting processing. Finally, the experimental results show that the proposed method obtains better performance than other methods in image stitching of outdoor scenes.