Detection and 3D Reconstruction of Buildings from Aerial Images
- Authors: Novotortsev L.V.1, Voloboy A.G.2
-
Affiliations:
- JSC Racurs
- Keldysh Institute of Applied Mathematics, Russian Academy of Sciences
- Issue: Vol 45, No 6 (2019)
- Pages: 311-318
- Section: Article
- URL: https://journals.rcsi.science/0361-7688/article/view/176903
- DOI: https://doi.org/10.1134/S0361768819060069
- ID: 176903
Cite item
Abstract
In this paper, we consider the problem of detection and 3D reconstruction of buildings from aerial images of the scene and usage of orientation data (camera parameters, GPS, etc.). Presently-available methods for solving this problem are inefficient on large amounts of data, where most of the data do not need thorough processing. In this work, we employ methods that find and analyze line segments on the image. This approach allows us to use image preprocessing based on line segment analysis, thereby reducing the amount of data that require computationally complex operations.
About the authors
L. V. Novotortsev
JSC Racurs
Author for correspondence.
Email: torets13@gmail.com
Russian Federation, ul. Yaroslavskaya 13А, Moscow, 129366
A. G. Voloboy
Keldysh Institute of Applied Mathematics, Russian Academy of Sciences
Author for correspondence.
Email: voloboy@gin.keldysh.ru
Russian Federation, Miusskaya pl. 4, Moscow, 125047