Region-based Satellite Image Classification: Method and Validation

Details

Ressource 1Download: BIB_552E48D7B010.P001.pdf (743.61 [Ko])
State: Public
Version: author
Serval ID
serval:BIB_552E48D7B010
Type
Inproceedings: an article in a conference proceedings.
Collection
Publications
Title
Region-based Satellite Image Classification: Method and Validation
Title of the conference
ICIP 2005, 12th IEEE International Conference on Image Processing
Author(s)
Gigandet X., Bach Cuadra M., Pointet A., Cammoun L., Caloz R., Thiran J.
Address
Genoa, Italy, September 11-14, 2005
ISBN
1522-4880
Publication state
Published
Issued date
2005
Series
Proceedings of the International Conference on Image Processing
Pages
III-832-III835
Language
english
Abstract
We propose an algorithm for very high-resolution
satellite image classification that combines
non-supervised segmentation with a supervised
classification. Both multi-spectral data and local
spatial priors are used in the Gaussian Hidden Markov
Random Field (GHMRF) model for the segmentation. Then,
two classifiers, Mahalanobis distance classifier and SVM,
are studied using intensity, texture and shape features.
Validation is done qualitatively and quantitatively by
comparison with a manual classification used as a ground
truth. Results show very good performance of our approach
in comparison to existing techniques. Also, we
demonstrate that spectral and spatial features calculated
on segmented regions are much more discriminant than the
spectral features of the pixels taken individually for
the classification task.
Keywords
bach, classification, Satellite imaging, segmentation
Create date
29/11/2011 16:40
Last modification date
20/08/2019 14:09
Usage data