Title :
Buildings extraction of very high spatial resolution satellite images
Author :
Benali, Abderraouf ; Dermeche, Hakima ; Belhadj, Sabrina ; Adnane, Akram ; Hanifi Elhachemi Amar, Reda
Author_Institution :
Dept. of Electron., LEPESA, Univ. of Sci. & Technol. of Oran (USTO-MB), Oran, Algeria
Abstract :
In this paper we propose a method based on mathematical morphology for a building detection of high spatial resolution urban areas satellite images, our method is unsupervised and does not require any additional data except the image itself. It consists on using several morphological operations such as Hit or Miss Transformation (HMT). The deformation correction and smoothing are used to eliminate the noise. Thereafter we compared our method with two other variants.
Keywords :
feature extraction; object detection; building detection; buildings extraction; deformation correction; deformation smoothing; hit or miss transformation; mathematical morphology; satellite images; Buildings; Image reconstruction; Morphology; Satellites; Smoothing methods; Spatial resolution; Transforms; Building extraction; HMT; Mathematical morphology; spatial resolution;
Conference_Titel :
Multimedia Computing and Systems (ICMCS), 2014 International Conference on
Conference_Location :
Marrakech
Print_ISBN :
978-1-4799-3823-0
DOI :
10.1109/ICMCS.2014.6911229
